TerraMosaic Daily Digest: May 6, 2026
Daily Summary
May 6's papers put rainfall-driven slope failure on stronger empirical and mechanistic ground. The Sancheong study combines a field-verified inventory, rainfall thresholds, GA-XGBoost, and SHAP to separate lithologic, topographic, and extreme-monsoon controls; the Shuikouyangou case shows how record rainfall can initiate debris flows in a watershed with no recognized debris-flow history through short-duration runoff concentration. Regional susceptibility papers from the Western Ghats, Rwanda, Greece, and the United States test how inventories, scale, climate forcing, and interpretable learning alter risk maps, while mechanics-oriented studies tie seismic anti-dip rock-slope failure, MT-InSAR creep, residual shear strength, root reinforcement, TRIGRS cluster failures, loess erosion, and seepage-driven particle migration to explicit failure processes.
Remote sensing is used less as a generic mapping layer and more as a measurement system with known failure modes. A spatiotemporal attentive U-Net denoises InSAR time series to recover millimetre-scale deformation; SAR change detection improves glacier displacement fields; SWOT-ICESat-2 differencing quantifies snow/ice penetration bias over glacial lakes; and terrain-aspect and strain priors make InSAR 3-D deformation inversion less ill-posed in mountains. The broader hazard set links mass-wasting sediment pulses through the Brahmaputra, thermokarst-lake storage response, wildfire burn probability, storm-surge dune erosion, Mozambique hazard-impact records, post-disaster recovery divergence, and infrastructure risk in open-pit mines, shield tunnels, landfills, utility tunnels, offshore foundations, and high-temperature underground works.
Key Trends
The strongest contributions convert event evidence into constraints on mechanisms, thresholds, and monitoring reliability.
- Rainfall thresholds are being linked to physical failure pathways: Sancheong landslides, Shuikouyangou debris flows, Jiangwan cluster failures, Heifangtai loess slopes, levee erosion, rooted soils, and unsaturated landfills all connect rainfall or seepage to measurable hydromechanical instability.
- Landslide susceptibility is widening from map production to scale and driver tests: Western Ghats, Rwanda, Greece, and United States studies compare conditioning factors, climate signals, inventory completeness, national-versus-regional scale, and model interpretability rather than treating AUC as the only endpoint.
- Deformation monitoring is becoming explicitly error-aware: InSAR denoising, MT-InSAR creep analysis, glacier pixel-offset tracking, 3-D deformation inversion, mining ground-displacement attribution, slow-slope seismic velocity monitoring, and SWOT lake-level correction all identify what the sensor misses before interpreting motion.
- Sediment and water redistribution dominate high-mountain hazard propagation: Brahmaputra mass-wasting pulses, glacier displacement, glacial-lake level bias, thermokarst-lake storage, and Tibetan Plateau water-storage attribution all connect upstream cryosphere or slope change to downstream hydrology and exposure.
- Hazard engineering is being linked to material behaviour, records, and institutional persistence: open-pit warning systems, shield-tunnel settlement, stockpile stability, tunnel support, utility-tunnel joints, subsea shield slurry, offshore foundations, historical earthquake evidence, and post-disaster recovery papers treat risk as a system that can lose reliability through material interfaces, design choices, missing records, or governance drift.
Selected Papers
This issue contains 59 selected papers from 1,670 papers analyzed. The leading papers resolve the Sancheong rainfall-induced landslide event, seismic anti-dip rock-slope stability, InSAR time-series denoising, open-pit landslide warning, Western Ghats susceptibility, MT-InSAR creep-type landslide dynamics, Rwanda climate-aware landslide mapping, ancient landslide reactivation, residual shear-strength controls on mobility, TRIGRS cluster landslide simulation, and extreme-rainfall debris-flow initiation. The broader set connects glacier displacement, glacial-lake water-level bias, Brahmaputra mass-wasting sediment routing, slow-slope seismic velocity monitoring, gully erosion, storm-surge dune erosion, mining ground fissures, rooted-soil mechanics, wildfire risk, historical earthquake evidence, landfill and stockpile stability, shield tunnel settlement, underground support, and remote-sensing methods for SAR soil moisture, land cover, dredge plumes, and infrastructure damage.
1. Physically based data-driven analysis for large-scale investigation of the July 2025 rainfall-induced landslide in Sancheong, South Korea
Core Problem: The July 2025 Sancheong landslide cluster requires event-scale separation of triggering rainfall, terrain controls, and material susceptibility.
Key Innovation: Remote sensing, field validation, rainfall threshold analysis, GA-optimized XGBoost, and SHAP interpretation are combined for 568 initiation points, showing critical EAR0 and cumulative rainfall thresholds for slope failure.
2. Stability evaluation method for anti-dip bedding rock slopes under seismic loading based on the Adam optimization algorithm
Core Problem: Anti-dip bedding rock slopes fail through flexural toppling and shear sliding, but pseudo-static failure-surface searches can be slow and unstable.
Key Innovation: A PS-Adam framework couples mechanical failure criteria with Adam optimization, matching shaking-table and numerical tests while greatly accelerating failure-surface search.
3. Denoising of InSAR Time Series Through Spatiotemporal Attentive Convolutional U‐Net
Core Problem: Atmospheric noise can conceal transient millimetre-scale deformation in InSAR time series.
Key Innovation: InSARDenoiser combines a spatial U-Net, temporal attention, and synthetic atmospheric noise training to recover slow deformation validated against creepmeter data on the North Anatolian Fault.
4. Multi-dimensional and multi-source disaster monitoring and warning system and case engineering application in Dagushan open-pit mining area
Core Problem: Open-pit mine slopes evolve under excavation and natural forcing, making monitoring-point layout and warning thresholds difficult.
Key Innovation: Statistical landslide analysis, damage-mechanics modelling, GNSS, InSAR, GIS, and mine IoT are fused into a multi-source warning platform for the Dagushan open-pit mine.
5. Landslide susceptibility mapping using a machine learning approach and different environmental factors in the Western Ghats region (India)
Core Problem: The Western Ghats need susceptibility products that account for rainfall, topography, geology, roads, land use, and seismicity together.
Key Innovation: Random Forest, SVM, stochastic gradient boosting, and bagging models are compared across 22 conditioning factors, identifying road proximity, deforestation, slope modification, and heavy rainfall as key controls.
6. An improved multi‐model risk assessment of creep‐type landslide dynamics with multi‐temporal synthetic aperture radar (MT‐InSAR)
Core Problem: Creep-type landslides have nonlinear hydroclimatic drivers and multiscale deformation that single-model assessments miss.
Key Innovation: Multi-orbit Sentinel-1 MT-InSAR, PS/DS processing, MGWR, K-medoids clustering, wavelet analysis, and hydroclimatic variables are integrated for risk assessment in Pakistan's Parsan Valley.
7. Deep learning models for comprehensive landslide susceptibility study considering climate change and sustainable development implications in Western Province of Rwanda
Core Problem: Mountainous Rwanda needs landslide-risk assessment that supports both disaster reduction and sustainable development planning.
Key Innovation: CNN, DNN, and MLP models are compared with SHAP analysis, identifying slope, DSM, and rainfall as dominant susceptibility controls in Western Province.
8. Study on deformation characteristics and failure modes of high-altitude ancient landslide in Lijie North Hill, Zhouqu, China
Core Problem: Ancient landslides along active fault zones can reactivate repeatedly and defeat existing control structures.
Key Innovation: Field investigation, drilling, laboratory tests, and time-series InSAR resolve deformation, rainfall response, and a three-stage failure mode for the Lijie North Hill landslide.
9. Experimental and numerical study of shear rate effects on residual shear strength and landslide mobility
Core Problem: Landslide mobility depends on shear-rate effects that differ by soil fabric and clay content.
Key Innovation: Ring shear tests and LS-Rapid simulations show opposing shear-rate effects at two Nilgiris landslide sites and quantify runout sensitivity to residual strength.
10. Dynamic simulation of slope stability based on TRIGRS model: a case study of cluster landslides in Jiangwan Town, Shaoguan, Guangdong, China
Core Problem: Southern China lacks field-parameterized regional simulations of shallow landslide stability during flood-season rainfall.
Key Innovation: TRIGRS is applied to the April 2024 Jiangwan cluster of 2,393 shallow landslides across dry, wet, and saturated states using field-derived parameters and observed rainfall.
11. Extreme rainfall and the initiation of debris flows: Insights from the Shuikouyangou case in Sichuan, China
Core Problem: Extreme rainfall can initiate debris flows in basins with no known debris-flow history.
Key Innovation: Field surveys, infiltration analysis, and hydrological simulations at Shuikouyangou link short-duration high-intensity runoff to rapid erosion and mobilization of loose material.
12. An Adaptive Homogeneous Pixel-Offset Tracking Technique Integrated With SAR Change Detection: A Case Study of Glacier Displacement Monitoring at Namcha Barwa in the Himalayas
Core Problem: Glacier displacement monitoring is degraded by SAR speckle noise, heterogeneous motion, and boundary decorrelation.
Key Innovation: DSTA-TransUNet change detection is coupled with adaptive homogeneous pixel-offset tracking to improve Namcha Barwa glacier displacement fields and completeness.
13. Enhancing gully erosion susceptibility prediction and agricultural land loss assessment through machine learning and remote sensing techniques
Core Problem: Gully erosion threatens agricultural land, but extraction accuracy and susceptibility controls remain uncertain.
Key Innovation: MNDWI, Fuzzy C-means, K-means, PCA, Boruta, and Fuzzy BWM are combined to map gully-affected areas and quantify 2000-2020 cultivated-land loss.
14. Numerical simulation of dune morphodynamics under storm surge: a case study of Hurricane Michael
Core Problem: Storm-surge erosion varies strongly alongshore, but the role of dune geometry needs event-scale simulation.
Key Innovation: XBeach simulations validated by post-Hurricane Michael field data identify crest-height thresholds, collision-regime dominance, and geometric controls on Mexico Beach dune erosion.
15. Landslide susceptibility assessment in Greece: work in regional and national scale
Core Problem: National susceptibility mapping must reconcile scale effects, official landslide inventories, and validation transparency.
Key Innovation: The updated Greek National Geodatabase of Landslides is used to compare statistical and expert-based susceptibility approaches across regional and national scales.
16. Tracking mass‐wasting sediments in the large Brahmaputra River system
Core Problem: Large mass-wasting events can perturb river systems far beyond the immediate source area, but downstream signals are rarely tracked basin-wide.
Key Innovation: Google Earth Engine turbidity, water-surface elevation, sandbar area, and flood extent reveal suspended-sediment propagation more than 1,000 km downstream in the Brahmaputra system.
17. Spatiotemporal trends of rainfall‑triggered landslides in the U.S. under climate change
Core Problem: National-scale links between changing rainfall and landslide activity are limited by uneven inventories.
Key Innovation: NASA COOLR data are analysed with emerging hot spot and colocation methods to identify spatiotemporal trends in U.S. rainfall-triggered landslides from 2006 to 2021.
18. A First Look at the Snow/Ice Penetration Effect of SWOT Observations on Water Level of Global Glacial Lakes
Core Problem: Glacial-lake water-level monitoring is limited by radar penetration through snow and ice cover.
Key Innovation: Near-synchronous SWOT and ICESat-2 observations over 260 global glacial lakes quantify a seasonal elevation bias and provide a correction framework for lake-variability records.
19. Decoupling Seasonal Seismic Velocity Changes on Slow‐Moving Slopes in Southwest China Using Tree Ensemble Machine Learning
Core Problem: Seismic velocity changes on slow-moving slopes reflect multiple meteorological and material controls.
Key Innovation: Two years of in situ seismic monitoring, XGBoost, and SHAP analysis decouple temperature, humidity, rainfall, air pressure, and local seismicity effects across rock and clastic-soil slopes.
20. Addressing Severe Class Imbalance in Spatio-Temporal Landslide Detection With Focal Tversky Loss
Core Problem: Landslides occupy tiny fractions of satellite time-series scenes, causing severe class imbalance and weak learning signals.
Key Innovation: Focal Tversky loss in a temporal-attention U-Net improves Sen12Landslides detection without adding DEM or SAR inputs.
21. Controlling mechanisms of rainfall intensity on erosion and progressive instability of unsaturated loess slopes in heifangtai: insights from integrated experimentation and coupled numerical simulation
Core Problem: Loess slopes can evolve from surface erosion to progressive instability under rainfall intensity changes.
Key Innovation: Integrated physical experiments and coupled numerical simulations quantify how rainfall intensity controls erosion and instability in Heifangtai loess slopes.
22. A UAV-Based Image Dataset and a Novel Morphological Feature Extraction Framework for Identifying Ground Fissures Caused by Coal Mining Activities
Core Problem: Mining-induced ground fissures are hazardous but lack public, high-resolution training data for segmentation.
Key Innovation: The GFM UAV dataset and FS-YOLO framework introduce fissure categories and morphology-aware instance segmentation for coal-mining disaster prevention.
23. Mechanical behaviors of rooted soil in triaxial tests and implications for rainfall-induced landslides
Core Problem: Vegetation can stabilize rainfall-prone slopes, but root geometry and biomass effects on shear behaviour remain uncertain.
Key Innovation: Triaxial tests on Chinese-fir rooted soils show how root biomass, angle, and diameter regulate shear strength, dilation, fracture, slippage, and static liquefaction resistance.
24. Burn probability and wildfire risk modeling of the protected areas in Golestan Province, NE Iran
Core Problem: Protected areas need spatially explicit burn probability and fire-risk maps for mitigation and climate adaptation.
Key Innovation: FlamMap minimum-travel-time simulations and support-vector regression generate burn-probability and risk estimates for protected areas in Golestan Province.
25. Micromechanics of Seepage Induced Fine Particle Migration in Soils with Different Grain Sequences: A Coupled CFD-DEM Analysis
Core Problem: Seepage can mobilize fine particles and weaken soils, but grain-sequence controls on migration are poorly constrained.
Key Innovation: Coupled CFD-DEM simulations compare positive and composite grain sequences under varying size ratios and hydraulic gradients, linking microstructure to migration velocity and displacement.
26. Experimental Investigation of Rainfall-Induced Erosion Control of River Levee Slopes Using Short Fiber Reinforcement
Core Problem: Levee slopes can lose soil rapidly through rilling and gully development during rainfall.
Key Innovation: Laboratory embankment rainfall tests quantify how coir and polypropylene short-fiber reinforcement delay rill formation and reduce eroded mass.
27. Multisource Data Integration Study on the Spatiotemporal Characteristics and Driving Mechanisms of Ground Displacement: Huainan City, China, Case Study
Core Problem: Ground displacement in mining cities reflects interacting groundwater, infrastructure, population, and economic drivers.
Key Innovation: SBAS-InSAR from 150 Sentinel-1 scenes is combined with GRACE-FO groundwater and socioeconomic layers using a geographical detector model for Huainan City.
28. Adaptive Fusion of Aspect Constraint and Strain Model for InSAR 3-D Deformation Inversion
Core Problem: Near-polar InSAR geometry is insensitive to north-south motion, making 3-D hazard deformation inversion ill-posed.
Key Innovation: Terrain-aspect constraints and strain models are fused into a unified InSAR line-of-sight inversion framework for high-resolution 3-D deformation.
29. An analytical method for predicting the ground surface settlement caused by shield tunnels with steep slopes and small-radius curves passing through existing buildings
Core Problem: Shield tunnels with steep slopes and small-radius curves create settlement patterns not captured by classical formulas.
Key Innovation: A Mindlin-solution analytical method is validated against Fuzhou Metro monitoring and evaluates thrust, friction, ground loss, building load, soil weight, and grouting pressure.
30. Statistical Analysis of Ionospheric TEC Anomalies Prior to Ms ≥ 6.0 Earthquakes in Mainland China During 2012–2022
Core Problem: Pre-earthquake ionospheric TEC anomalies remain statistically contested and require systematic event-scale characterization.
Key Innovation: Mainland China Ms >= 6.0 earthquakes from 2012 to 2022 are analysed by anomaly polarity, magnitude, focal depth, azimuth, and timing before events.
31. Natural hazard impacts database for Mozambique using information from 1926 to 2021: knowing the past to plan for a resilient future
Core Problem: Resilience planning needs local, process-specific records of hazard impacts beyond global databases.
Key Innovation: Newspaper, official, and NGO records from 1926 to 2021 are harmonized at district scale and compared with DesInventar and EM-DAT.
32. Coupled flow–deformation and slope stability analysis of MSW landfills on unsaturated collapsible loess: A case study
Core Problem: Leachate infiltration and groundwater depth can weaken collapsible loess beneath landfills under static and seismic conditions.
Key Innovation: Coupled seepage, stress-deformation, transport, and slope-stability modelling quantifies leachate-induced strength degradation at the Azadshahr landfill.
33. Storage–supply regulation and rainfall responses driven by thermokarst lakes in permafrost-degrading catchments
Core Problem: Thermokarst lakes alter rainfall-runoff coupling, groundwater response, and storage release in degrading permafrost catchments.
Key Innovation: Paired catchments on the Qinghai-Tibet Plateau are analysed using water-balance closure, event diagnostics, groundwater response, lake-stage change, and hysteresis metrics.
34. Numerical modeling of the out-of-plane dynamic response of masonry gable walls via a high-fidelity block-based finite element modeling approach – part II: post-diction and application to other structural configurations
Core Problem: Masonry gable walls remain vulnerable to earthquake out-of-plane failure and require calibrated high-fidelity models.
Key Innovation: Block-based finite-element models are recalibrated after blind-prediction experiments and extended to structural configurations beyond the original tests.
35. Constitutive model and thermal-induced damage behaviors of dolomite subjected to water–rock interaction
Core Problem: Environmentally eroded carbonate rocks show nonlinear compaction that many damage models omit.
Key Innovation: A segmented damage constitutive model couples thermal-wet-cycle water-rock interaction with load-induced damage and validates pore and skeleton deterioration by SEM, NMR, and solution chemistry.
36. Multiscale Testing and Application Research on Structured Hybrid Alkali-Resistant Glass Fiber-reinforced Shotcrete for Ultra-High Ground Temperature Tunnels
Core Problem: Tunnel primary support must retain integrity under high ground temperature, high stress, and rockburst-prone conditions.
Key Innovation: Microscopic, mechanical, and field tests show that hybrid alkali-resistant glass fibers suppress cracking and reduce support stresses in an under-construction highland railway tunnel.
37. A preliminary design method for jacket support structures of high-capacity offshore wind turbines
Core Problem: Offshore wind foundations must be preliminarily sized under coupled aerodynamic, hydrodynamic, pile, and natural-frequency constraints.
Key Innovation: A layer-specific jacket design workflow and public MATLAB tool rapidly size members and verify frequency and deflection constraints.
38. A hybrid experimental/numerical method to predict traffic-induced environmental vibrations for ultra-sensitive scientific facilities
Core Problem: New scientific facilities near traffic infrastructure require reliable prediction of construction-stage and operation-stage vibrations.
Key Innovation: A hybrid experimental-numerical workflow inverts virtual moving loads from field ground responses and reconstructs wavefields with multi-objective calibration.
39. Diagnosing Long-Term Divergence in Build Back Better Outcomes in Low- and Lower-Middle Income Communities
Core Problem: Recovery investments can drift away from hazard-informed design and zoning objectives over time.
Key Innovation: Housing assessments, surveys, interviews, and stakeholder consultations in Banda Aceh and Central Sulawesi identify five pathways by which Build Back Better outcomes erode.
40. Quantifying the disproportionate climate risks of Indigenous People and Afro-Descendants in rural Honduras: using transport accessibility to basic services as a new metric for equity-centered climate risks
Core Problem: Climate-risk assessments can miss how transport disruption isolates vulnerable rural communities from health and education services.
Key Innovation: A Honduras case study uses accessibility to basic services as an equity-centered metric for disproportionate climate and natural-hazard risk.
41. Probabilistic Slope Stability Analyses of an Unsaturated Heterogenous Stockpile of Rockfill and Granular Soil Mixture
Core Problem: Stockpile slopes near friction-angle limits can become unstable when precipitation or snowmelt alters apparent cohesion.
Key Innovation: Limit-equilibrium and Monte Carlo analyses quantify how unsaturated shear-strength variability and geometry affect stockpile stability.
42. Research on Road Surface Distress Detection Algorithm in UAV Images with Multi-Scale Feature Fusion
Core Problem: Pavement distress detection from high-angle UAV images suffers from small target scale and feature aliasing.
Key Innovation: An improved YOLOv8 with feature-focusing diffusion pyramid fusion, interaction heads, and edge extraction improves small-distress recognition for infrastructure inspection.
43. Comprehensive Evaluation of the GF-3 Series SAR Satellites for Soil Moisture and Surface Roughness Retrieval over Bare Soils
Core Problem: Large-scale SAR soil-moisture retrieval must be validated across roughness, texture, topography, and climate gradients.
Key Innovation: GF-3 SAR retrieval strategies are evaluated against 378 quality-controlled field plots across 11 arid-to-humid regions in China.
44. Understanding the Optical Behavior and Spectral Signature of Dredging-Induced Plumes in Coastal Waters
Core Problem: Dredge plumes can perturb coastal waters, but their optical signature must be separable from natural turbidity.
Key Innovation: Laboratory samples and satellite water reflectance reveal a flatter absorption spectrum that supports Sentinel-2 monitoring of dredging plumes.
45. Artificial Intelligence and Cloud Computing for a New Generation of Corine Land Cover Maps in Colombia
Core Problem: Disaster and environmental planning require up-to-date land-cover information at finer resolution than legacy expert interpretation.
Key Innovation: AI and cloud workflows generate CORINE-compatible 1:50,000 land-cover maps for Colombia with semi-automated annual update potential.
46. A hybrid recurrence–wavelet transform approach for geochemical classification of rocks along boreholes
Core Problem: Mine and tunnel design needs rapid, consistent interpretation of multivariate drillhole assay data.
Key Innovation: Recurrence analysis, wavelet transforms, and machine learning automatically delineate geochemical domains and transitional boundaries along boreholes.
47. Mechanical performance of L-shaped joints in utility tunnels with Large-Spacing reinforcement
Core Problem: Reinforcement congestion can weaken utility-tunnel joint construction and integrity.
Key Innovation: Full-scale static tests and analytical stress-transfer modelling evaluate large-spacing reinforcement in double-compartment utility tunnel joints.
48. Kinetic equivalence-based testing and DEM analysis of the muck transfer mechanism in inclined shaft TBM
Core Problem: Inclined shaft TBMs face low muck-transfer efficiency and chute wear under gravity-driven transport.
Key Innovation: A kinetic-equivalence test platform and DEM model simulate irregular muck transfer from cutterhead to chute and quantify design controls on efficiency.
49. Rheological properties and underlying mechanisms of a novel eco-friendly slurry for saline environments
Core Problem: Seawater ionic strength complicates slurry selection for subsea shield tunnelling.
Key Innovation: Sugarcane-bagasse-modified bentonite slurry is optimized and microstructurally evaluated for viscosity, suspension stability, and sealing performance in seawater.
50. Heatwave characterization in Chile according to the Excess Heat Factor (1979–2018)
Core Problem: South American heatwave hazard remains under-characterized from a biometeorological perspective.
Key Innovation: High-resolution gridded temperature data and synoptic analysis characterize Chilean heatwave intensity, frequency, duration, extent, trends, and representative extreme-event patterns.
51. Previously Unknown Historical Evidence from Parish Registers of Irpinia Earthquakes (Southern Italy) During the Modern Age
Core Problem: Historical earthquake risk assessment depends on sparse, uneven, and often unpublished impact records.
Key Innovation: Parish registers from Irpinia provide previously unknown casualty evidence, support revised macroseismic interpretations for seventeenth- and eighteenth-century earthquakes, and document post-event construction responses.
52. A Method to Implement Natural Flow Regimes for Regulated Rivers
Core Problem: Dam operations must balance flood-control and allocation constraints with ecological flow variability.
Key Innovation: An optimization framework synchronizes dam releases with local hydrology while preserving water-allocation and flood-control requirements.
53. Probabilistic Agro‐Hydrology: A Stochastic Framework for Irrigation Risk Assessment and Water Management
Core Problem: Deterministic irrigation models miss rare combinations of dry spells and peak crop-water demand.
Key Innovation: A stochastic agro-hydrological framework estimates irrigation return periods and water-demand risk across climatically diverse Italian regions.
54. Seismic performance of various precast concrete beam-to-column joints: A review and test data-based investigation
Core Problem: Precast frame joints vary widely in connection detail and earthquake resistance.
Key Innovation: A test-data-based review classifies precast beam-column joints and compares strength, stiffness, deformation capacity, and energy dissipation.
55. Full-scale tests and theoretical analyses of responses for combined post-grouted drilled shafts with various slenderness ratios and bearing strata
Core Problem: Foundation settlement prediction depends on slenderness ratio, grouting, and bearing strata parameters.
Key Innovation: Full-scale load tests and theoretical analysis produce response-curve predictions for combined post-grouted drilled shafts.
56. Attributing terrestrial water storage changes on the Tibetan Plateau to climate and human drivers using a hybrid deep learning approach
Core Problem: Tibetan Plateau water-storage dynamics are hard to attribute because of satellite gaps and interacting climate and human drivers.
Key Innovation: A VMD-CNN-LSTM framework reconstructs and attributes terrestrial-water-storage change, separating precipitation, evapotranspiration, runoff, temperature, and residual components.
57. Functional zonation of enzyme-mediated biogeochemical processes along lateral hydrological connectivity gradients in river floodplains
Core Problem: Floodplain function varies with lateral hydrological connectivity, but enzyme-mediated spatial zonation remains unclear.
Key Innovation: Soil enzyme sampling and partial least squares path modelling identify functional zones across high-, intermediate-, and low-connectivity river-floodplain settings.
58. Role of Crop Salt Tolerance in Enhancing Remote Sensing-Based Soil Salinity Mapping Across Irrigated Agroecosystems: A Review
Core Problem: Soil-salinity mapping remains surface-oriented and often ignores crop tolerance and root-zone dynamics.
Key Innovation: A structured review of 58 studies shows how crop salt tolerance mediates soil-plant-spectral relationships and affects transferability of salinity maps.
59. Quantitative analysis of the mechanical behavior of granular sands subjected to particle pre-crushing
Core Problem: Stress history and particle crushing control granular material behaviour in foundations and transport earthworks.
Key Innovation: Drained triaxial tests and acoustic-emission monitoring quantify how pre-crushing shifts energy dissipation from crushing-dominated to frictional rearrangement.