TerraMosaic Daily Digest: May 22, 2026
Daily Summary
May 22 is unusually strong in slope and cryosphere-linked hazard mechanics. The highest-value landslide contribution reconstructs the 2025 Blatten rock-ice avalanche as a domino-style cascade in which rock collapse destabilized glacier ice, generated a high-velocity mixed avalanche, produced air blasts, and formed a barrier lake. Two additional high-impact papers broaden the mechanics of geohazard forcing: a Science analysis shows that earthquake-related rock damage can weaken landscapes tens of kilometers from active faults, while a Science Advances study demonstrates that paleo-salt water, not only modern seawater intrusion, dominates salinization in many subsiding Chinese coastal aquifers.
The day's landslide and hydrologic papers are methodologically concrete. Rainfall-induced landslide mapping couples slope-stability modelling with machine learning; DDA-SPH resolves coupled seepage and block motion in soil-rock mixture slopes; microseismic monitoring exposes internal fracture evolution in high mine slopes; and full-scale debris-flow measurements constrain impact forces for mitigation design. Hydrologic and climate-hazard work links process models, remote sensing, and AI across permafrost sediment connectivity, urban storm modification, flood-prone area mapping, typhoon ocean response, Arctic streamflow, extreme precipitation scaling, and drought-related water storage. Remote-sensing papers add physically interpretable observables: 3D mining displacement from SAR, hydrologic loading deformation from InSAR, postfire recovery trajectories, SAR-optical snow cover, mine-tailings heavy metals, and efficient segmentation for disaster response.
Key Trends
The day is organized around coupled failure chains and measurable state variables: rock-ice avalanche progression, fault-damaged rock strength, rainfall-driven slope failure, debris-flow impact force, levee runout, permafrost sediment connectivity, urban storm response, coastal salinity, flood-index mapping, typhoon wave and ocean structure, and SAR/InSAR-derived deformation.
- Landslide and mass-movement studies emphasize coupled cascades: the Blatten rock-ice avalanche reconstruction, rainfall-landslide hybrid model, DDA-SPH soil-rock slope simulation, microseismic mine-slope monitoring, active-waveguide AE experiments, root-reinforced excavation tests, and LLM-based urban landslide reconstruction treat slope failure as linked triggering, internal damage, geometry, and runout.
- Earthquake and rock-damage papers connect rupture processes to landscape and infrastructure response: fault-zone rock damage, Hikurangi slow slip, locked strike-slip analog models, tunnel shaking in fault-fracture zones, active-fault tunnel analysis, seismic fragility modelling, and abandoned-mine full-waveform inversion translate seismic or mining processes into deformation and stability metrics.
- Hydrologic hazards are framed through dynamic connectivity and extremes: permafrost sediment connectivity, delta response to extreme flows, GFI 2.0, IDF regionalization, extreme-precipitation scaling, typhoon wave assimilation, Arctic streamflow prediction, and snowmelt-DOC mapping focus on when and where water, sediment, and solutes become connected.
- Coastal and groundwater studies shift from simple boundary intrusion to subsurface memory and heterogeneity: coastal salinization, offshore groundwater salinity, DNAPL inversion, preferential-flow AI, nitrate and trace-organic groundwater risk, tidal arsenic cycling, and beach-aquifer nitrogen dynamics show the control of aquifer architecture, paleo conditions, and tidal exchange.
- Remote sensing contributes most when it returns hazard-relevant physical quantities: InSAR elastic loading, LuTan-1 3D mining displacement, postfire NDVI reconstruction, SAR-optical snow-cover fusion, UAV hyperspectral mine-tailings mapping, snow BRDF modelling, and LightFormer segmentation improve deformation, water storage, fire recovery, snow, contamination, and disaster-response mapping.
Selected Papers
This issue contains 63 selected papers from 1,260 papers analyzed. The leading papers center on coupled geohazard mechanics: the 2025 Blatten rock-ice avalanche cascade, global tectonic rock damage and erosion, continental-scale paleo-salt-water control on coastal aquifer salinization, rainfall-induced landslide mapping with physics and machine learning, DDA-SPH rainfall slope failure, microseismic mine-slope instability, full-scale debris-flow impact forces, and probabilistic seepage-driven levee failure and runout. The broader set follows water, sediment, fire, ice, and infrastructure hazards through permafrost sediment connectivity, urban storm response, flood-prone area indexing, typhoon ocean and wave observations, Arctic streamflow, postfire vegetation recovery, 3D mining deformation, SAR-optical snow mapping, frozen-soil and subgrade degradation, tunnel-fault seismic response, landslide acoustic emission, root-reinforced slopes, abandoned-mine imaging, aquifer contamination, and lightweight remote-sensing segmentation for disaster response.
1. A domino-style rock-ice avalanche at Blatten, the Swiss Alps, in 2025
Core Problem: High-mountain rock-ice avalanches can amplify through coupled rock collapse, glacier destabilization, air blast, and lake formation, but full event sequences are rarely reconstructed.
Key Innovation: Satellite imagery, seismic records, numerical simulations, and eyewitness videos reconstruct the 2025 Blatten event as a domino-style cascade from rock collapse to mixed rock-ice avalanche and downstream inundation.
2. Role of tectonic rock damage in erosional processes: A global analysis
Core Problem: The spatial reach of earthquake-induced rock weakening has been uncertain, limiting how tectonic damage is represented in erosion and hillslope-hazard models.
Key Innovation: A global Science analysis of cosmogenic erosion rates shows that elevated erosional efficiency can extend tens of kilometers from active faults, linking seismic damage to landscape-scale rock-strength reduction.
3. Paleo–salt water dominates coastal aquifer salinization: A continental-scale study in China
Core Problem: Coastal aquifer salinity is often attributed to modern seawater intrusion, but the role of trapped paleo-salt water has remained poorly quantified at continental scale.
Key Innovation: A Science Advances study of more than 2100 samples across China shows that paleo-salt water dominates many salinized sedimentary basins and is strongly associated with tectonic subsidence zones.
4. An enhanced framework for mapping rainfall-induced landslides by integrating physics-based numerical modelling with machine learning
Core Problem: Real-time rainfall-induced landslide mapping must combine mechanistic slope stability with data-driven performance under extreme storms.
Key Innovation: A Hong Kong framework integrates a slope-stability model, 24 years of rainstorm-landslide records, and machine learning, improving validation against the June 2008 storm beyond standalone physics or ML models.
5. Development of a DDA-SPH Method for Hydromechanical Modeling of Rainfall-Induced Instability in Soil-Rock Mixture Slopes
Core Problem: Soil-rock mixture slopes require models that simultaneously resolve block motion, unsaturated seepage, and strong hydromechanical coupling during rainfall.
Key Innovation: A coupled DDA-SPH method simulates rainfall-induced instability by combining discrete rock-block mechanics with SPH seepage-deformation modelling for large-deformation slope failure.
6. Microseismic monitoring-based stability analysis of high rock slopes in deep open-pit mines
Core Problem: Surface displacement alone can miss internal fracture evolution in deep open-pit mine slopes approaching instability.
Key Innovation: A dense microseismic array, displacement monitoring, geology, and numerical simulation reconstruct a 3.5-year progressive creep failure in the Changshanhao open-pit mine slope.
7. Measurements and analysis of natural debris flow impacts in the Gadria Creek
Core Problem: Full-scale debris-flow impact forces remain difficult to constrain, especially when boulder impacts and bulk-flow forces overlap.
Key Innovation: In-situ measurements at Gadria Creek use load cells, radar, ultrasonic sensors, and video to separate surge forces from single-boulder impacts and quantify design-relevant forces.
8. Probabilistic assessment of seepage-induced levee failure and post-failure runout using a GPU-accelerated SPH–random field framework
Core Problem: Levee risk depends on spatially variable materials, progressive seepage failure, and post-failure spreading, which are costly to sample probabilistically.
Key Innovation: A GPU-accelerated SPH-random-field framework couples soil and pore-water phases with anisotropic random fields to simulate failure initiation, large deformation, and runout across Monte Carlo scenarios.
9. Cryo‐FSC: A Dynamic Functional Connectivity Index Reveals Spatio‐Temporal Shifts in Sediment Transport in the Permafrost‐Dominated River Systems
Core Problem: Cold-region sediment connectivity indices often fail to capture seasonal freeze-thaw and snowmelt controls on sediment flux.
Key Innovation: Cryo-FSC integrates cryospheric and terrain-process variables to quantify dynamic hillslope-channel coupling in the Source Region of the Yangtze River and tracks connectivity shifts from 2000 to 2022.
10. Divergent urban storm response to convective, frontal and tropical systems
Core Problem: Urban rainfall studies have reported inconsistent effects because storm type and scale are often mixed together.
Key Innovation: A Nature analysis of more than 40,000 Texas storms classifies storm types and shows that cities affect local, frontal, and tropical systems differently in occurrence and intensity.
11. Unlocking the potential of InSAR elastic loading deformation for geodetic joint inversion of terrestrial water storage changes in California
Core Problem: InSAR deformation caused by hydrologic elastic loading is underused in joint inversions of terrestrial water storage.
Key Innovation: A Remote Sensing of Environment study incorporates time-series InSAR with GRACE/GRACE-FO and GNSS to improve California water-storage inversion during the 2019-2022 drought.
12. Geomorphic flood index 2.0: enhanced tools for delineating flood-prone areas in data-scarce regions
Core Problem: DEM-based flood indices can misrepresent inundation near confluences where tributary-main stem interactions change local water levels.
Key Innovation: GFI 2.0 adds an iterative confluence module guided by stream hierarchy and improves flood-prone mapping across Italian basins without full hydraulic modelling.
13. High-resolution dataset of 2024 typhoons in the northern South China Sea captured by a collaborative network of underwater gliders and autonomous underwater vehicles
Core Problem: Traditional typhoon ocean observations are too sparse to resolve three-dimensional upper-ocean evolution during storm passage.
Key Innovation: An ESSD dataset provides high-resolution temperature, salinity, and current observations from underwater gliders and autonomous vehicles during 2024 northern South China Sea typhoons.
14. Geodetic Resolution of the Interplay Between Earthquakes and Slow Slip in the Hikurangi Margin
Core Problem: The temporal relationship between deep slow slip and moderate earthquakes remains difficult to resolve from sparse inversions.
Key Innovation: Three independent geodetic inversion strategies applied to six years of GNSS data recover a consistent 2021-2024 Hikurangi slow-slip history and its interaction with Mw 4-5 earthquakes.
15. Non‐Linear Morphodynamic Response of Deltas to Extreme Events
Core Problem: Deltas exposed to changing extreme-flow frequency may respond nonlinearly, complicating flood and coastal-landform management.
Key Innovation: pyDeltaRCM simulations show that delta size, channel number, width, and slope vary nonlinearly across extreme-flow intermittency regimes, yielding a conceptual model of delta adjustment.
16. Freeze–Thaw Controls on Dissolved Organic Carbon Export: A Numerical Investigation of Hydrological Mechanisms in a Hillslope
Core Problem: Permafrost carbon budgets need mechanisms linking seasonal freeze-thaw dynamics to dissolved organic carbon mobilization.
Key Innovation: A process-based hydro-thermal-geochemical hillslope model in the Qinghai-Tibet Plateau resolves seasonal groundwater flowpaths controlling DOC concentration and export timing.
17. A hybrid Kolmogorov–Arnold Networks-based model with attention for predicting Arctic river streamflow
Core Problem: Arctic river discharge is difficult to predict because of sparse observations, strong seasonality, and cryosphere-hydrology coupling.
Key Innovation: A hybrid Kolmogorov-Arnold Networks, LSTM, and attention model with seasonal encoding and physical constraints predicts Arctic streamflow across rapidly changing basins.
18. Uncovering the melt: Unmanned Aircraft System (UAS) and in-situ sensor synergies reveal dissolved organic carbon (DOC) pathways in a northern peatland
Core Problem: DOC export during spring snowmelt depends on fine-scale snow depletion and hydrologic connectivity that are hard to map with conventional sensors.
Key Innovation: UAS snow-depth mapping, topographic wetness, and high-frequency stream monitoring reveal when snow-free wet zones initiate DOC leaching in a northern peatland.
19. Introducing Graph Attention Network to Improve Wang Zhengfei Semi-Empirical Model for Accurately Simulating Forest Fire Spread in Western Guizhou Province
Core Problem: Cellular-automata fire-spread models can be too rigid to represent nonlinear propagation across heterogeneous terrain and fuels.
Key Innovation: A graph attention network adjusts CA neighborhood rules using historical Guizhou fire incidents and validates simulated spread against Sentinel-2 burned patches.
20. The Fire Modeling Intercomparison Project (FireMIP) for CMIP7
Core Problem: Global fire models need coordinated evaluation inside coupled Earth-system models as fire extremes intensify.
Key Innovation: FireMIP for CMIP7 defines experiments to evaluate past, present, and future fire regimes in coupled models and assess fire feedbacks on vegetation, carbon, and society.
21. From reanalysis to climatology: deep learning reconstruction of tropical cyclogenesis in the western North Pacific
Core Problem: Tropical cyclogenesis climatology is difficult to reconstruct because environmental drivers are high-dimensional and observational labels are uncertain.
Key Innovation: TCG-Net uses an 18-layer residual CNN and tailored labelling strategies to reconstruct western North Pacific cyclogenesis from reanalysis at temporal and spatial prediction tasks.
22. Regionalization of IDF curves for mainland China: a comparative evaluation of machine learning versus spatial interpolation techniques
Core Problem: Sub-daily intensity-duration-frequency curves are essential for drainage design but difficult to regionalize where observations are sparse.
Key Innovation: A mainland China comparison evaluates station-based interpolation and gridded-precipitation machine learning methods for IDF curves from 1 to 72 hours and 2 to 1000-year return periods.
23. Global assessment of extreme precipitation in response to climate warming: revisiting the Clausius–Clapeyron relation
Core Problem: Observed precipitation-temperature scaling can diverge from Clausius-Clapeyron expectations depending on how extremes are defined.
Key Innovation: A global analysis introduces an unconditional extreme-precipitation approach that evaluates intensity changes at temperatures where true extremes occur.
24. Assimilation of multi-source satellite wave observations into the COAWST model during typhoons using deep learning
Core Problem: COAWST lacks a dedicated wave-assimilation module, limiting representation of high sea states during typhoons.
Key Innovation: A 3D CNN U-Net assimilates satellite altimeter significant-wave-height observations into COAWST and improves spatial wave fields for independent 2025 typhoon cases.
25. Modeling wave-induced sediment transport and morphodynamic response in a bar-built estuary
Core Problem: Wave-dominated estuaries can reorganize sediment pathways during storms, especially under higher water levels.
Key Innovation: Phase-resolving FUNWAVE-TVD simulations quantify infragravity-wave controls on suspended and bedload transport in an idealized Chilean bar-built estuary.
26. Postfire Vegetation Recovery Assessment Through Time Series Reconstruction of Vegetation Index
Core Problem: Clouds, atmospheric noise, and phenological variability obscure consistent postfire recovery trajectories.
Key Innovation: An NDVI time-series reconstruction framework based on double-logistic fitting and iterative optimization retrieves annual phenological metrics for heterogeneous postfire recovery assessment.
27. Integrating earth observation data into the tri-environmental evaluation of the economic cost of natural disasters: a case study of 2025 LA wildfire
Core Problem: Wildland-urban interface fires require fine-grained assessment of dynamic impacts across population, built assets, and ecosystems.
Key Innovation: A 2025 Los Angeles wildfire case study integrates VIIRS daily fire detections, OpenStreetMap, high-resolution population modelling, and tri-environmental indicators to track evolving impacts in Eaton and Palisades.
28. Quantifying post-fire differences in forest structure and aboveground biomass using the TECIS full-waveform LiDAR system
Core Problem: Large-scale post-fire assessment is limited by sparse observations of three-dimensional forest structure and biomass change.
Key Innovation: A hierarchical framework links field plots, airborne LiDAR, TECIS full-waveform LiDAR footprints, and Sentinel-1/2 imagery to quantify post-fire structure and biomass differences.
29. Derivation of 3-D Surface Displacements in Mining Areas Using Farneback Optical Flow Method With LuTan-1 Ascending and Descending SAR Imagery
Core Problem: Large-gradient mining deformation can break InSAR coherence while conventional offset tracking remains limited in accuracy and efficiency.
Key Innovation: A Farneback optical-flow method with LuTan-1 ascending and descending SAR imagery resolves vertical, east-west, and north-south displacement components in mining areas.
30. XFuse: Multi-sensor CNN transformer fusion with cross attention and quality aware curriculum learning for high resolution fractional snow cover mapping
Core Problem: Fractional snow-cover retrieval requires fusing optical spectral information with all-weather SAR while handling resolution mismatch.
Key Innovation: XFuse combines CNN-transformer architecture, bidirectional cross-attention, and quality-aware curriculum learning for high-resolution SAR-optical fractional snow-cover mapping.
31. Prediction of dynamic parameters and construction of dynamic constitutive parameter model for permafrost based on machine learning
Core Problem: Cold-region infrastructure design needs frozen-soil dynamic parameters across temperature, water content, confining pressure, and loading frequency.
Key Innovation: Dynamic triaxial tests and interpretable machine-learning models predict frozen silty-sandstone elastic modulus and construct a constitutive parameter evolution model.
32. Mechanism of ice melting in silty clay subgrade driven by hydrothermal coupling: A case study of the G0615 Dekang Expressway in Eastern Qinghai-Tibet Plateau
Core Problem: Water infiltration can accelerate ice melting in plateau road subgrades, but the hydrothermal coupling sequence is poorly resolved.
Key Innovation: A Qinghai-Tibet Plateau case study identifies four stages of silty-clay subgrade ice melting and shows how infiltration velocity reshapes heat transfer and low-temperature zones.
33. Application of an artificial horizontal aquiclude with slurry diaphragm walls for groundwater control in deep excavations near high-speed railways
Core Problem: Deep excavations near high-speed railways can induce unacceptable settlement when dewatering thick confined aquifers.
Key Innovation: An artificial horizontal aquiclude integrated with suspended slurry diaphragm walls and recharge control isolates groundwater while limiting settlement-sensitive impacts.
34. Energy-based analysis of seismic interaction between tunnel lining and surrounding rock in fault-fracture zones: A shaking table test study
Core Problem: Mountain tunnels crossing fractured fault zones face severe seismic damage from heterogeneous surrounding rock and lining interaction.
Key Innovation: A large-scale shaking-table test of the Zheduoshan Highway Tunnel quantifies energy transfer, damage evolution, and lining-rock interaction under seismic loading.
35. Seismic analysis of tunnels crossing active faults using an improved stochastic ground motion simulation method
Core Problem: Tunnels crossing active faults need ground-motion inputs that reflect source, path, and local uncertainty.
Key Innovation: An improved stochastic ground-motion simulation method is used to analyze seismic response of tunnels crossing active faults.
36. Experimental study on the shear behaviour and acoustic emission of active waveguide under soil–structure interaction
Core Problem: AE waveguides can support landslide early warning, but signal-generation mechanisms under shearing remain insufficiently understood.
Key Innovation: Refined shear experiments show three-phase active-waveguide deformation and synchronous AE responses controlled by particle size, shear speed, and normal stress.
37. Centrifuge modeling of plant roots reinforcement on soil slope under excavation conditions
Core Problem: Root reinforcement can alter deformation and slip surfaces during excavation, but the mechanics are difficult to observe at prototype scale.
Key Innovation: Centrifuge tests and 3D-printed roots show that root reinforcement deepens failure mechanisms, reduces displacement locally, and raises critical excavation height.
38. Reconstruction of landslide events in urban setting using LLM-based Agentic AI with multimodal data
Core Problem: Urban landslide investigations require synthesizing witness reports, images, engineering evidence, and geometry estimates quickly after events.
Key Innovation: An LLM-based agentic AI workflow combines RAG extraction, YOLOv8 segmentation, and multimodal prompting to reconstruct landslide dimensions and debris volumes from historical events.
39. Local stress evolution in locked strike-slip fault analog models: Fault activity characterization
Core Problem: The local stress evolution around locked strike-slip segments controls instability but is hard to observe experimentally.
Key Innovation: Analog fault models with embedded strain cubes and DIC track principal stress rotation and deformation around locked segments of varying strength.
40. Delineating abandoned mines using full-waveform inversion of shallow-seismic narrow land-data streamer in Galena, Kansas
Core Problem: Abandoned mining cavities pose surface-stability risks, but urban near-surface complexity makes void delineation difficult.
Key Innovation: A two-stage elastic full-waveform inversion of shallow land-streamer seismic data maps abandoned mine cavities beneath Galena, Kansas using velocity, density, and surface-wave information.
41. Mechanism and Computational Method for Phased Deformation of Deep Loess during Water Level Rise
Core Problem: Deep loess can deform in phases during water-level rise, affecting road and foundation stability.
Key Innovation: The study develops a mechanism and computational method for phased deformation of deep loess under rising water levels.
42. Monotonicity-constrained ordinal Gaussian process for seismic fragility modeling
Core Problem: Bridge fragility models need flexible nonparametric prediction while avoiding physically impossible curve crossing.
Key Innovation: A monotonicity-constrained ordinal Gaussian-process framework improves Wenchuan earthquake bridge damage prediction and enforces ordered fragility behaviour.
43. Characterizing Multi‐Source DNAPL Source Zone Architecture in Channelized Aquifers Using a Geophysics‐Informed Hybrid Deep Learning Framework
Core Problem: Multi-source DNAPL contamination in channelized aquifers is difficult to invert because permeability fields, leakage locations, and saturation are coupled.
Key Innovation: A GAN-CNN-geophysics inversion framework assimilates electrical-resistivity observations to reconstruct channelized permeability, source locations, leakage rates, and DNAPL saturation.
44. Insights from nuclear magnetic resonance investigations in tailings storage facilities
Core Problem: Tailings engineering has limited shallow-subsurface NMR field evidence for estimating water content and layering in storage facilities.
Key Innovation: NMR investigations at three tailings storage facilities are interpreted with CPTu, pore-pressure dissipation, seismic velocity, resistivity, and laboratory data to assess applicability and limits.
45. The Role of Link Length Hierarchy on Dynamic Connectivity in River Networks Across Climatic Regimes
Core Problem: Basin-scale transport depends on how river-network geometry and topology control dynamic connectivity across climates.
Key Innovation: A 100-basin U.S. analysis separates link-length hierarchy from branching topology and shows how climate shapes flux aggregation rates.
46. Deep Learning‐Based Particle Tracking Velocimetry (PTV) for Spherical and Non‐Spherical Particles: Application to Sediment Transport
Core Problem: Dense, rotating, non-spherical particles limit conventional particle tracking in sediment-transport experiments.
Key Innovation: YOLO-based object detection and Kalman filtering recover particle-scale velocity, rotation, shear-rate, and granular-temperature profiles for idealized and natural particles.
47. Do Deep‐Learning Models Perform Better Than Process‐Based Models? A Diagnostic Evaluation Using Synthetic Streamflow Simulation Under Various Sources of Uncertainties
Core Problem: Daily streamflow models are sensitive to measurement, precipitation, and out-of-sample uncertainty in ways that are rarely tested systematically.
Key Innovation: Synthetic controlled experiments compare LSTM, Sacramento process model, and hybrid approaches under known uncertainty sources.
48. Accuracy and Limitations of Tools for Analyzing Meander Migration: Consequences of Bend Dynamics and Analysis Interval
Core Problem: Centerline-based meander migration tools can produce biased trajectories when bend translation is rapid or analysis intervals are poorly chosen.
Key Innovation: Numerical and real-river tests evaluate MTMT and dynamic time warping, showing accuracy limits tied to bend dynamics and migration rate.
49. An Upper Bound on Carbon Emissions of Drained Peat Soil Grasslands From Satellite Radar Interferometry
Core Problem: Regional carbon-loss estimates from drained peatlands lack spatially resolved empirical constraints.
Key Innovation: Satellite radar interferometry separates reversible and irreversible parcel-scale land motion to estimate an upper bound on CO2-equivalent emissions and evaluate subsidence mitigation.
50. Effects of climate change on desert dust
Core Problem: Climate effects on the desert dust cycle remain uncertain because aridity, vegetation, snow cover, lake levels, wind, and sediment supply interact nonlinearly.
Key Innovation: A review synthesizes how warming, drought, dryland hydrology, vegetation, biocrusts, and atmospheric variability may increase or reduce future dust emissions.
51. On the calculation of the land cover and management factor in soil erosion assessments
Core Problem: USLE/RUSLE/MUSLE erosion estimates remain sensitive to heterogeneous methods for calculating the land cover and management factor.
Key Innovation: A review classifies C-factor approaches and compares major methods in the Olt River Basin to quantify their assumptions and implications.
52. Traditional water meadows (lameiros) buffer the increase in sediment transport caused by recurrent fires in Mediterranean headwater basins
Core Problem: Recurrent fires can raise sediment connectivity in headwater basins, but valley-bottom water meadows may modulate the response.
Key Innovation: A paired-basin analysis uses burn severity, sediment connectivity, and sediment delivery ratios to show how lameiros buffer post-fire sediment transport.
53. Hydroclimatic forcing and phase-locked soil erosion in the Weihe Basin since late Middle Pleistocene based on geochemical and sedimentary evidence
Core Problem: Long records of erosion intensity are needed to understand how hydroclimate phases control landscape denudation.
Key Innovation: Geochemical and sedimentary evidence reconstructs six late Middle Pleistocene erosion stages in the Weihe Basin and links them to hydroclimatic transitions.
54. The visible and hidden climatic effects on Earth's denudation
Core Problem: The influence of climate on physical and chemical denudation varies with terrain slope and aridity.
Key Innovation: A global model of 10Be-derived denudation rates separates visible water-biological effects from hidden slope-mediated climate effects on erosion.
55. Spatiotemporal dynamics of land degradation in the three gorges reservoir area and its coupling mechanisms with biodiversity and human activity intensity
Core Problem: Long-term ecological stability in the Three Gorges Reservoir Area requires integrated assessment of degradation, biodiversity, and human activity.
Key Innovation: A 1990-2024 multi-source geospatial model combines vegetation, salinization, bare rock, soil fertility, water erosion, productivity, and land-use indicators.
56. Study on the friction properties of compacted snow layers and influencing factors
Core Problem: Snow-layer friction depends on temperature, contamination, density, thickness, and surface topography but is poorly parameterized.
Key Innovation: Field and controlled laboratory tests quantify friction coefficients of compacted and wind-eroded snow layers under varying environmental and material conditions.
57. Critical state concept for rock joints: Description of dilatancy hardening
Core Problem: Rock-joint dilatancy under constant normal stiffness requires constitutive descriptions that couple compression and shear.
Key Innovation: A critical-state framework calibrated with numerical simulations and direct shear data models normal-stress and shear-stress evolution during constrained rock-joint shearing.
58. Paleo-hydrological controls on continental-shelf groundwater salinity distribution offshore New Jersey, USA
Core Problem: Offshore fresh groundwater persistence depends on paleo sea-level history and density-driven transport, but salinity controls are uncertain.
Key Innovation: Seismic, borehole, geochemical, isotope, and 3D variable-density flow modelling reconstruct offshore New Jersey groundwater emplacement during Pleistocene lowstands.
59. A multi-AI framework for accurate characterization of preferential flow pathways using a field tracer experiment
Core Problem: Non-Gaussian preferential flow pathways are difficult to infer efficiently from field tracer data.
Key Innovation: A multi-AI inversion framework combines spectral-normalized WGAN priors, data assimilation, and groundwater flow-solute transport models to characterize high-conductivity pathways.
60. Arsenic cycling driven by hydrogeochemical-tidal coupling in a sandy subterranean estuary
Core Problem: Submarine groundwater discharge can transport arsenic to coastal waters, but tidal recirculation and aquifer redox zonation complicate flux estimates.
Key Innovation: High-resolution field observations show how recirculated saline groundwater dominates seaward arsenic flux while intertidal sediments act as both filter and conduit.
61. Unraveling dissolved inorganic nitrogen dynamics in a sandy beach: The role of tidal forces and aquifer heterogeneity
Core Problem: Nitrogen fluxes from intertidal aquifers are uncertain because redox zonation and tidal cycles vary with sediment heterogeneity.
Key Innovation: Spring-neap monitoring reveals vertically stratified nitrate and ammonium zones and shows how aquifer heterogeneity controls dissolved inorganic nitrogen production and transport.
62. A UAV Hyperspectral Inversion Framework for Mapping Soil Heavy Metals Based on Spectral Harmonization, Weighted Ensemble Learning, and Environmental Variable Integration
Core Problem: Heavy-metal contamination in mine tailings is spatially discontinuous and difficult to map with sparse laboratory sampling alone.
Key Innovation: A UAV-laboratory hyperspectral framework harmonizes spectra and uses weighted ensemble learning to map Cd and Pb in an open-pit mine with interpretable variable importance.
63. LightFormer: An efficient lightweight decoder for optical remote sensing semantic segmentation
Core Problem: Disaster-response segmentation requires efficient decoders that can handle unstructured targets without excessive computational cost.
Key Innovation: LightFormer uses lightweight feature fusion, channel processing, learnable gating, and spatial refinement to improve optical remote-sensing semantic segmentation for time-sensitive applications.