TerraMosaic Daily Digest: July 17, 2026
Daily Summary
Landslide forecasting is moving from isolated displacement traces to spatially resolved deformation fields. A ground-based radar index that combines cumulative deformation, moving-area extent and coherence forecasts two contrasting mine-slope failures within 4-6 minutes and provides 36 minutes of warning where single-point analysis fails. At the regional scale, spatial validation of nine susceptibility models in Hiroshima shows that similar conventional metrics can conceal large differences in where high-hazard classes are assigned, while a constrained polar-angle formulation represents irregular critical slip surfaces without admitting kinematically implausible geometries.
A small change in material strength can be amplified through an entire hazard cascade. Temperature-dependent ring-shear tests and process-chain modelling of the 2021 Melamchi disaster show that a friction-angle reduction below 0.6 degrees produces longer runout, larger landslide dams, higher debris-flow peaks and greater impact pressure under a late-century warming scenario. This mechanism links climate forcing to geomaterial response rather than treating warming as an external statistical covariate.
Across other hazards, the strongest advances separate physical change from observational or statistical artefact. Sediment and glacial isostasy contribute 0.7-1.5 mm per year of subsidence near the Ayeyarwady mouth; a nonstationary compound-flood framework isolates the distinct upstream and estuarine effects of rainfall, surge and antecedent flow; and sparse early ocean profiles are shown to create a spurious upward trend in marine-heatwave intensity. Aftershock-conditioned fragility, near-fault railway risk, probabilistic Vs30 inversion and InSAR-mechanics integration similarly make uncertainty part of the hazard estimate rather than an unreported residual.
Key Trends
Five linked shifts connect spatial warning, cascade mechanics, scale dependence and explicit uncertainty.
- Spatial fields replace single-point warning signals: Radar coherence, deformation-zone extent and multi-temporal InSAR resolve where motion grows and which physical mechanism controls it.
- Cascade models propagate material sensitivity: Temperature-dependent shear strength is carried through runout, damming, breach discharge and impact pressure instead of stopping at slope initiation.
- Risk changes with scale and nonstationarity: Settlement-landslide coupling, compound flooding and climate extremes yield different controls across spatial bands, time windows and urban settings.
- Observation support becomes an explicit uncertainty source: Profile density, dispersion-curve uncertainty, sparse labels and LiDAR geolocation errors are quantified before physical trends or mapped classes are interpreted.
- Sequential loading enters consequence models: Mainshock-aftershock pairs, pulse-like near-fault motion and underdrainage-driven soil-structure interaction replace isolated-event or greenfield assumptions.
Selected Papers
The leading papers connect deformation sensing, slope mechanics and process-chain propagation to outcomes that can be tested: warning time, mapped hazard, runout, flood depth, structural loss and quantified uncertainty.
1. An improved landslide early warning and forecasting method based on multi-component ground-based radar data
Core Problem: Single-point displacement series discard most of the spatial information in ground-based radar and can miss composite landslides that do not follow classical creep acceleration.
Key Innovation: The EOCD index fuses cumulative deformation, moving-area extent and coherence into a spatial warning signal, forecasting two contrasting failures within 4-6 minutes and providing 36 minutes of warning where point analysis failed.
2. Changing post-landslide cascade hazard intensity due to increasing temperature: a process-based simulation of the 2021 Melamchi disaster in Nepal
Core Problem: Climate-warming effects on landslide cascades are commonly imposed through rainfall scenarios while temperature-dependent changes in geomaterial strength remain unresolved.
Key Innovation: Ring-shear constraints and process-chain modelling show that a friction-angle reduction below 0.6 degrees can lengthen runout, enlarge landslide dams and increase debris-flow height, sediment flux and infrastructure impact pressure.
3. Present Estimations of Subsidence Induced by Sediment Isostasy and Glacial Isostatic Adjustment in the Ayeyarwady Delta (Myanmar)
Core Problem: The contribution of sediment loading and post-glacial mass redistribution to present subsidence had not been quantified for the densely populated Ayeyarwady Delta.
Key Innovation: A viscoelastic Earth model constrained by stratigraphic sedimentation histories estimates 0.7-1.5 mm per year of isostatic subsidence near the river mouth, adding a distinct component to relative sea-level and flood-risk budgets.
4. Decoding Compound Flood Risk Under Nonstationary Conditions: A Separable Parametric Framework Revealing the Heterogeneous Effects of Driving Mechanisms
Core Problem: Compound-flood studies often predict inundation without isolating how rainfall, storm surge and antecedent streamflow produce spatially changing risk under nonstationarity.
Key Innovation: A separable parametric framework couples an inundation estimator with driver-conditioned exceedance rates, reproducing Greater Boston storms while distinguishing rainfall-led upstream change from estuarine modulation near the coast.
5. A Systematic Comparison of Statistical and Machine-Learning Models for Mapping Landslide Susceptibility: Evidence from the 2018 Rainfall-Induced Landslides in Hiroshima
Core Problem: Landslide-susceptibility models can appear equivalent under aggregate classification metrics while assigning observed failures to very different spatial hazard classes.
Key Innovation: Nine statistical and machine-learning models are tested on a balanced 2018 Hiroshima inventory; spatial validation places 96.72% of landslides in high classes for weights of evidence versus 72.54% for SVM despite comparable conventional performance.
6. A novel framework for generalized potential slip surfaces in complex slopes using polar angle incrementation
Core Problem: Regular slip surfaces cannot represent complex failures, whereas unconstrained random surfaces can be kinematically inadmissible and computationally inefficient.
Key Innovation: A physically constrained polar-angle increment function, coupled to a tensile-shear criterion and genetic search, generates flexible critical surfaces with few parameters and reproduces complex slope mechanisms in numerical and engineering cases.
7. Scale dependency: Re-examining the underlying linkages between the evolution of rural settlement patterns and landslide risk in hilly and mountainous regions
Core Problem: Single-scale spatial analysis obscures how settlement expansion and contraction interact with landslide distributions at different characteristic lengths.
Key Innovation: Variational mode decomposition and wavelet coherence separate dominant settlement and landslide scales of about 7.7 km and 1.2 km and reveal localized coupling below 8 km that changes between the 1995-2010 expansion and 2010-2025 contraction phases.
8. Dynamic response and seismic fragility analysis of a bridge abutment located on flat and sloping grounds designed using code-based procedure
Core Problem: Seismic design methods for retaining systems do not directly capture the coupled effects of slope geometry, superstructure inertia and out-of-phase earth pressure on bridge abutments.
Key Innovation: Plane-strain analysis compares sloping and flat sites, benchmarks Mononobe-Okabe and Newmark estimates, and derives fragility curves that quantify how slope placement changes abutment demand and displacement.
9. Vector-valued fragility surfaces conditioned on mainshock-aftershock intensity levels for masonry buildings
Core Problem: Single-mainshock fragility models cannot represent cumulative damage when aftershocks arrive before masonry buildings can be inspected or repaired.
Key Innovation: Sequential nonlinear analyses of physically consistent mainshock-aftershock pairs produce bivariate fragility surfaces directly from both intensity levels and reveal substantial loss amplification even after moderate mainshocks.
10. Traffic risk assessment method for high-speed railway bridge networks considering the spatial and probabilistic characteristics of pulse-like ground motions
Core Problem: Railway-network risk models can misstate near-fault disruption by ignoring the spatial occurrence probability and amplification of pulse-like ground motions.
Key Innovation: A source-to-site probabilistic framework propagates pulse effects from bridge-level function loss to network connectivity and safe speed, showing 25-70% risk error when their spatial distribution is omitted.
11. Fine-scale deformation pattern and mechanism of high-fill canals revealed by multi-temporal InSAR
Core Problem: High-fill canal deformation combines long-term consolidation with seasonal soil response, but conventional InSAR interpretation does not separate these superimposed mechanisms.
Key Innovation: Distributed-scatterer SBAS analysis of Sentinel-1 and TerraSAR-X data, interpreted with soil mechanics, distinguishes prolonged upper-canal consolidation from precipitation-driven shrink-swell motion in lower sections.
12. Incorporating the impact of floating piles in estimation of large-scale subsidence
Core Problem: Large-scale underdrainage damage models commonly apply greenfield settlement to piled buildings and therefore neglect protective soil-structure interaction.
Key Innovation: Time-dependent finite-element modification factors reduce predicted long-term pile settlement by about 20% and change the low-damage classification for roughly 23% of piled buildings in a Gothenburg demonstration.
13. Seismic retrofit design optimization of connection-head-gapped braces using reduced-order model
Core Problem: The dynamic impact behaviour of connection-head-gapped seismic braces is too costly to resolve repeatedly during retrofit optimization.
Key Innovation: A reduced-order impact model coupled to metaheuristic design controls drift in 10- and 20-storey structures while retaining acceptable acceleration and force demands.
14. Quantifying Vs30 uncertainty using dispersion curves derived from surface wave methods
Core Problem: A single inverted shear-wave profile masks the epistemic and random uncertainty in Vs30 estimates derived from measured dispersion curves.
Key Innovation: A constrained Monte Carlo workflow propagates dispersion-curve uncertainty through profile selection and recovers both the mean and standard deviation of Vs30 with relative errors below 5% in synthetic tests.
15. A fast method for calculating topographic envelope surfaces based on a reverse topography drainage area threshold
Core Problem: Topographic-envelope reconstruction becomes computationally expensive when high-resolution DEMs require windows wider than deeply incised valleys.
Key Innovation: Reverse-topography drainage areas identify crest points without large moving windows, yielding a fast envelope method that remains effective on degraded-resolution DEMs across canyon and mature landscapes.
16. Fracture-Matrix Flow Partitioning Governs Retention Regime Transitions in Unsaturated Fractured Rock
Core Problem: Unsaturated fractured-rock models struggle to upscale the sharp hydraulic contrast between connected fractures and the porous matrix.
Key Innovation: Direct three-dimensional simulations and an analytical retention law identify a critical saturation at which flow switches between matrix- and fracture-dominated regimes, with a closed-form limit for connected networks.
17. Prediction of uniaxial compressive strength from drilling parameters using metaheuristic-optimized ensemble regression models
Core Problem: Direct rock-strength testing is slow and destructive, limiting rapid characterization during drilling.
Key Innovation: A 345-sample stacking ensemble optimized by particle swarm search predicts uniaxial compressive strength from penetration rate, rotation, torque and thrust, with penetration and rotation emerging as the dominant controls.
18. Mechanical behavior and critical state constitutive model of granite soil-rock mixture subjected to drying-wetting cycles
Core Problem: Drying-wetting degradation of granite soil-rock mixtures is not represented adequately by constitutive models calibrated to a single hydraulic state.
Key Innovation: Triaxial tests show five cycles reduce shear strength by 15.6-20.1% and shear modulus by 41.7-57.5%; a state-dependent critical-state model reproduces the associated contraction-to-dilation transition.
19. Impact of rainfall intensity on soil moisture dynamics and wetting front migration: insights from one-dimensional soil column experiments
Core Problem: Wetting-front formulas that assume complete saturation do not reproduce changing matric suction under intense or intermittent rainfall.
Key Innovation: A soil-column model resolves downward wetting and bottom-up saturation and predicts front migration without the full-saturation assumption, with maximum errors of 6% against experiments and 8.5% against simulations.
20. LC-SLab - An object-based deep learning framework for large-scale land cover classification from satellite imagery and sparse in-situ labels
Core Problem: Sparse in situ labels produce fragmented large-area land-cover predictions when pixel-wise deep learning is applied without a spatial minimum mapping unit.
Key Innovation: LC-SLab combines graph-based input aggregation or object-level output aggregation with pretrained features, matching or exceeding pixel baselines while producing more coherent Sentinel-2 maps under sparse supervision.
21. GEDICorrect: A Scalable Python Tool for Orbit-, Beam-, and Footprint-Level GEDI Geolocation Correction
Core Problem: Residual orbit bias and pointing jitter degrade GEDI terrain and canopy metrics at footprint scale.
Key Innovation: The open-source GEDICorrect framework performs orbit-, beam- and footprint-level waveform matching, improving canopy-height R2 from 0.61 to 0.78 and scaling to 24 cores with about 19.5-fold runtime reduction.
22. Insights into the mechanism of settlement and stability of cantilever retaining walls under tunnel-induced ground movements using finite element analysis
Core Problem: Tunnel-induced settlement and lateral loading on nearby cantilever walls depend jointly on tunnel depth, volume loss and excavation direction.
Key Innovation: Centrifuge-validated three-dimensional analysis shows advancement direction can increase wall rotation by 62% and shallow tunnelling can raise lateral force by 230%, while resolving settlement, sliding and overturning together.
23. Effect analysis of twin tunnels crossing karst cavities beneath an elevated structure based on complex geological modeling
Core Problem: Horizontally layered ground models can underestimate tunnel-induced deformation where karst cavities and fractured strata are spatially uncertain.
Key Innovation: A coupled Markov-chain geological model feeds probabilistic cavity geometry into FLAC3D; field comparisons show improved deformation prediction and quantify how face pressure changes structural risk.
24. Predicting the (active) width of gravel-bed braided rivers
Core Problem: Width predictors calibrated for single-thread channels systematically underrepresent the morphology of gravel-bed braided rivers.
Key Innovation: Laboratory and New Zealand field data show that conventional bankfull-width equations predict the active, not total, braided width and support a minimum corridor near three times the predicted bankfull width.
25. Spatially Heterogeneous Impacts of Urbanization in Reshaping the Dry and Hot Extremes Across Chinese Cities
Core Problem: National assessments rarely separate urbanization from background climate forcing in compound hot-dry extremes.
Key Innovation: Multi-decadal observations across China attribute 2-20% of heat intensification since the 1970s, and 12-46% during 2003-2022, to urbanization while linking urban growth to reduced light rainfall and stronger inland drought.
26. Temporal Heterogeneity of In Situ Ocean Observing Capacity Could Cause an Artificial Intensification of Extreme Warm Water Events Globally
Core Problem: The apparent increase in marine-heatwave intensity from in situ profiles is an order of magnitude larger than trends in reanalysis and climate simulations.
Key Innovation: The discrepancy is traced to the transition from sparse to dense profile coverage: early sampling underestimates extremes and creates an artificial trend exceeding 10% per decade, requiring observation-density correction before attribution.
27. Sedimentation and Nonlinear Trapping in Texas Reservoirs Identified Using Remote Sensing and Bathymetric Survey Records
Core Problem: Sparse bathymetric surveys leave reservoir-storage loss poorly constrained between expensive field campaigns.
Key Innovation: Satellite sediment-concentration estimates, calibrated against repeated Texas bathymetry, identify high-loss reservoirs and recover nonlinear trapping patterns that can support more frequent capacity assessment.
28. Propagation and statistical distribution characteristics of storm waves under the shelter of the tombolo: A case study at the nearshore of Chudao, China
Core Problem: Tombolo sheltering and tidal stage alter nearshore storm-wave propagation, but the resulting changes in wave-height tails and nonlinearity are poorly constrained by field data.
Key Innovation: Measurements from two storms show that Weibull statistics fit open-water waves, whereas the Glukhovskiy distribution better resolves the sheltered high-wave tail; diffraction also creates a bimodal height-period distribution and stronger low-tide attenuation.
29. Preserving Spatial Fidelity in Heterogeneous Landscapes: A Biophysical Index-Guided 1 × 1 CNN for Multi-Source Remote Sensing Fusion
Core Problem: Coarse ecological products blur abrupt boundaries in fragmented landscapes, while neighbourhood-based downscaling can import values across physically distinct surfaces.
Key Innovation: A biophysical-index-guided 1 x 1 CNN reconstructs productivity without spatial mixing; although a 3 x 3 CNN is slightly better globally, the point-wise model preserves gradients and lowers full-domain spatial error over a mined landscape.
30. A Fast Adjacency Effect Correction Algorithm for High-Spatial-Resolution Optical Satellite Imagery with Adaptive Local Surface Constraints
Core Problem: Adjacency scattering degrades high-resolution optical imagery, while conventional local-background corrections become inaccurate in heterogeneous scenes and costly over large influence windows.
Key Innovation: Adaptive surface constraints and finite-window FFT convolution reduce GaoFen-2 reflectance error below 0.006 and process a 1024 x 1024 image with a 501-pixel window in under 4 seconds.
31. Non-Iterative Autofocus Method for High-Resolution SAR Time-Domain Imaging Based on Multi-Subimage 2D-PGA
Core Problem: Iterative alternation between SAR imaging and motion compensation is too computationally demanding for rapid ultra-high-resolution time-domain imaging.
Key Innovation: Multi-subimage 2D phase-gradient autofocus estimates and splices spatially varying wavenumber errors within parallel ground-Cartesian back-projection, validated on measured 0.03 m Ku-band SAR data.
32. Multi-Resolution Estimation of GNSS-Based Terrestrial Water Storage Changes Using Green’s Function and Slepian Basis Function Methods
Core Problem: Nominal grid refinement can overstate the effective resolution of GNSS-based terrestrial-water-storage inversions when station density and hydrological coherence are inadequate.
Key Innovation: Green's-function and Slepian inversions of 2,839 GNSS stations at 1, 0.5 and 0.25 degrees show that finer grids often amplify noise; effective resolution instead depends on network density, signal coherence and regularization.
33. Research on Few-Shot Mars Rover Onboard Surface Scene Classification Based on SE-ResNet-MTL
Core Problem: Onboard surface-scene classifiers must adapt from very few labelled images without sacrificing performance when a fuller training set is available.
Key Innovation: An annotated Mastcam benchmark and SE-ResNet meta-transfer model reach 95.52% validation accuracy and 90.5% and 91.5% in one- and five-shot tasks, improving on conventional fine-tuning by 42.8% and 25.4%.
34. Mechanical behavior, failure characteristics and energy evolution of layered rock subjected to uniaxial loading
Core Problem: Loading rate and lithologic layering jointly control energy transfer and fracture localization, but their cross-scale effects are not resolved by single-material tests.
Key Innovation: Coupled compression, acoustic-emission, digital-image-correlation and particle-flow analyses identify four failure stages and show that a sandy-mudstone middle layer increases simulated crack counts by 40-75%.
35. Mapping participatory approaches in disaster risk management: A scoping review
Core Problem: Participation is promoted throughout disaster-risk management, but its depth and influence across mitigation, preparedness, response and recovery remain unevenly documented.
Key Innovation: A scoping review of 51 studies finds co-creation and co-action concentrated in mitigation and preparedness, with response and recovery constrained by power asymmetries, limited resources, uneven representation and weak institutional uptake of local knowledge.
36. Urban heat and public green space: Assessing risk, equity, and the lived experiences of residents in Bochum, Germany
Core Problem: Urban-heat maps can misdirect adaptation when cooling access, spatial inequality and residents' lived exposure are evaluated separately.
Key Innovation: A Bochum assessment combines heat risk, public-green-space cooling and resident experience to identify tracts where high exposure coincides with weak cooling access and persistent inequity.
37. Spectrally-informed parametric graph DeepONet for response prediction on unstructured meshes of aerospace structures
Core Problem: Fixed-mesh surrogates lose accuracy when structural geometry, finite-element discretization and excitation spectra vary together.
Key Innovation: A spectrum-conditioned graph DeepONet learns full-field responses on unstructured meshes and reaches an R2 of 0.958 with 5.7% relative L2 error for acceleration on unseen geometries.
38. On simple baselines for domain shift in condition monitoring: A case study in bearing fault identification
Core Problem: Domain-adaptation claims are difficult to compare when datasets confound changes in operating conditions, fault severity, preprocessing and task definition.
Key Innovation: More than 600,000 controlled evaluations and a Relative Performance Drop metric show that rotational-speed shifts cause 42.8% degradation, whereas load shifts are negligible; the accompanying taxonomy distinguishes solvable pipeline shifts from ill-posed tasks.
39. Fréchet-derivative-based global sensitivity analysis for models with correlated inputs
Core Problem: Global sensitivity measures based on output-distribution shape generally assume independent inputs, despite correlated parameters in physical models.
Key Innovation: A Nataf-copula, kernel-density and change-of-measure formulation extends Frechet-derivative sensitivity to correlated non-Gaussian inputs and approaches 100,000-1,000,000-sample Monte Carlo accuracy with hundreds to roughly one thousand evaluations.
40. Spatiotemporal bias in OCO-2 XCO2 retrievals relative to TCCON observations
Core Problem: Long-term satellite carbon-dioxide retrievals can carry season-, land-cover- and viewing-mode-dependent biases relative to ground observations.
Key Innovation: A ten-year comparison across 17 TCCON sites identifies a latitude-dependent OCO-2 nadir bias, with land-cover associations reaching R2 values of 0.95 above 45 degrees N and 0.79 between 0 and 45 degrees N.
41. Flow dynamics in an erodible triangular channel bank cavity
Core Problem: Riverbank-cavity studies based on rigid rectangular forms do not represent flow separation, scour and deposition around erodible triangular cavities.
Key Innovation: Flume measurements resolve a 68.3% velocity reduction in the cavity and a 63.4% rise in shear stress along its shear layer, linking recirculation-zone deposition and downstream scour to bank-stabilization design.
42. An unconstrained implicit stress integration algorithm for the HS model
Core Problem: Three-dimensional implementation of the Hardening Soil model is hindered by non-smooth yield-surface corners, principal-stress permutations and repeated active-set switching.
Key Innovation: Smooth constitutive reconstruction and a CHKS complementarity function recast stress integration as one unconstrained Newton system; an ABAQUS UMAT matches laboratory and Plaxis references and remains robust under localized footing plasticity.
43. Adaptive step-size update and mixed-residual stopping criterion for EG/CI-type projection algorithms in DDA-d frictionless contact analysis
Core Problem: Projection solvers in dual discontinuous-deformation analysis advance slowly early in contact iteration and may stop late without directly testing non-penetration and complementarity.
Key Innovation: A delta-driven step-size rule and mixed contact-residual criterion reduce iteration cost and local peaks while improving terminal constraint satisfaction and contact-force accuracy without changing the underlying projection-correction scheme.
44. Coral reef and mangrove interplay during Holocene sea level changes, Belitung Island, Indonesia
Core Problem: Coral reefs and mangroves are often projected separately even though their long-term response to sea-level change depends on shared accommodation and shoreline history.
Key Innovation: Outcrop stratigraphy and sediment cores show a shift from vertical reef accretion to horizontal progradation around 4.3-3.8 ka, creating broad platforms that subsequently accommodated terrestrial sediment and mangrove expansion.
45. Quantifying Groundwater Recharge With Cosmogenic Isotopes and a Hybrid Age-Ranked Storage Selection Model in California (USA)
Core Problem: Recharge estimates must resolve strong year-to-year precipitation variability, but groundwater age models are often weakly constrained for young water.
Key Innovation: A hybrid age-ranked storage model combines sulfur-35, tritium and tritiogenic helium-3, reducing transit-time uncertainty by 27% on average and resolving sharply different recharge contributions from wet 2023 and drier 2024 conditions.
46. Quantifying Dynamic River Gains and Losses Using Inverse Water Temperature Modeling
Core Problem: Direct measurements of river gains and losses are expensive and provide limited temporal or spatial coverage of surface-water and groundwater exchange.
Key Innovation: Inverse modelling of inexpensive temperature-logger data reconstructs exchanges over five months; fine segmentation best recovers gain-loss direction, while intermediate segmentation best matches total exchanged volume.
47. Scale effects on span elongation beneath pipelines and cables in live-bed conditions
Core Problem: Small laboratory models may not reproduce the rate at which live-bed scour lengthens unsupported spans beneath subsea pipelines and cables.
Key Innovation: Tests at 5, 10 and 20 cm diameters isolate pronounced scale effects for the smallest deeply embedded model and relate declining span-growth rates to ripple height relative to exposed structure diameter.
48. Comment on Guth et al. Benchmarking Elevation Plus Land Surface Parameters Finds FathomDEM and Copernicus DEM Win as Best Global DEMs. Remote Sens. 2025, 17, 3919
Core Problem: A widely cited global-DEM ranking lacked significance tests, reusable inputs and explicit correction for pixel misalignment.
Key Innovation: A reproducible reanalysis adds Kruskal-Wallis testing and pixel alignment, confirms FathomDEM's leading rank but finds no significant overall quality differences and no barren-terrain degradation of GEDTM v1.2 relative to Copernicus DEM.
49. Active Earth Pressure on Counterweight Walls Under Different Movement Modes
Core Problem: Active earth pressure on counterweight walls changes with translation, toe rotation, heel rotation and foundation bearing capacity, complicating identification of the governing design state.
Key Innovation: A two-slip-surface analytical formulation minimizes sliding, overturning and bearing-capacity safety factors for each movement mode, agrees with experiments and simulations to about 15%, and identifies translation and toe rotation as the more readily mobilized states.
50. Air flow dynamics through a breached foredune and relationships to trough blowout formation: A Computational Fluid Dynamics (CFD) study
Core Problem: Flow acceleration through natural dune breaches and artificial notches is poorly constrained across breach and foredune geometries.
Key Innovation: Computational experiments show that taller, steeper foredunes and narrower breaches strengthen and extend high-speed jets; increasing foredune height from 4 to 7 m shifts jet termination from 20 to 45 m downwind.
51. Three-phase sediment coarsening in the Lower Yangtze River reveals downstream response to Tibetan Plateau uplift since ~3 Ma
Core Problem: The lower Yangtze lacks a continuous sedimentary record linking downstream grain-size change to late-Pliocene Tibetan Plateau uplift.
Key Innovation: Five-centimetre-resolution gamma and grain-size records identify coarsening at about 2.8-2.6, 1.8-1.7 and 1.1-1.0 Ma, intervals better aligned with uplift episodes than with monsoon or global-climate variability.
52. Quantitative analysis of crust and loose layer characteristics on the surface of the earthen Great Wall
Core Problem: Environmental erosion creates a surface crust over a loose subsurface layer on the earthen Great Wall, but the particle-scale controls on this loss of integrity remain poorly quantified.
Key Innovation: Multifractal particle-size analysis, Gaussian-mixture clustering and PLS-SEM separate western and eastern controls, quantify direct stability pathways, and show biocrust-associated gains of 28% and 49% in aggregate-size metrics.
53. Impacts of deforestation on soil physical and hydraulic properties in the hyrcanian forest, Iran
Core Problem: Deforestation effects on soil structure and water retention vary with both land-use history and depth, which surface-only comparisons cannot separate.
Key Innovation: Profiles to 1 m across six land uses show 14-19% topsoil porosity loss under agriculture, persistent structural degradation after orchard conversion and improved deep-water retention in sandier orchard soils.
54. Multiscale investigation of hydrate pore habit effects on the mechanical properties of hydrate-bearing sediments
Core Problem: Sediment strength depends on whether gas hydrate fills pores, cements contacts or carries skeletal load, yet models rarely represent all three habits together.
Key Innovation: A hybrid DEM and DEM-FEM framework reproduces triaxial tests across hydrate saturation and ranks cementing hydrate as the strongest and most saturation-sensitive control on stiffness, peak strength and shear-band development.
55. Establishing the payment for the ecosystem service of soil erosion control
Core Problem: Soil-erosion payments based on mean rainfall erosivity may underfund conservation where intense storms drive losses above a defensible tolerance.
Key Innovation: A Sicily-wide framework links payment targets to USLE crop-management, support-practice and slope-length factors, showing when land-cover change alone is insufficient and where combined interventions exceed the target payment.
56. Evolution and mechanisms of tidal duration asymmetry in the Xijiang River Channel of the Pearl River Estuary
Core Problem: Flood- and ebb-duration imbalance in estuaries responds nonlinearly to river discharge, tidal range and engineered shoreline change.
Key Innovation: Fifty years of water levels and Delft3D-FM simulations reveal a discharge threshold of 11,000-16,000 m3/s between flood and ebb dominance and attribute a step toward ebb dominance to reclamation-driven amplification of the M2 tide.