TerraMosaic Daily Digest: Feb 7, 2026
Daily Summary
This digest synthesizes 43 selected papers and focuses on groundwater and subsurface hydrogeologic hazards, landslide process mechanics and slope evolution, wildfire hazard dynamics and adaptation. Top-ranked studies examine risk, fragility, and resilience assessment, wildfire hazard and adaptation, and groundwater and subsurface hydrogeologic hazards.
Across the full set, evidence converges on mechanism-constrained analysis with operational relevance, especially for high-resolution remote-sensing monitoring workflows and freeze-thaw and cryosphere-driven instability. The strongest contributions pair interpretable process evidence with monitoring or forecasting workflows that support warning design and risk prioritization.
Key Trends
- Hydrogeologic hazards are treated through coupled subsurface dynamics: Groundwater variability, transport processes, and deformation signals are linked to monitoring and management decisions.
- Landslide studies increasingly resolve process chains: Contributions connect triggering conditions, slope deformation, and mobility outcomes, improving the basis for warning thresholds and scenario testing.
- Wildfire research is integrated with broader geohazard management: Physical drivers, landscape controls, and operational planning are analyzed together to evaluate cascade risk.
- Monitoring workflows rely on integrated remote-sensing products: Multi-source satellite and airborne observations are used for deformation retrieval, change detection, and rapid post-event mapping.
- Cryosphere and freeze-thaw effects remain first-order controls: Studies quantify thaw-related weakening and cold-region instability relevant to hazard evolution and design.
Selected Papers
This digest features 43 selected papers from 574 papers analyzed (out of 574 raw papers scanned) across multiple journals. Each paper has been evaluated for its relevance to landslide and broader geohazard research and includes links to the original publications.
1. Integrating Physical Drivers for Wildfire Hazard Modelling in Support of Disaster Risk Reduction: A Case Study in Castile and León, Spain
Core Problem: The urgent need for updated, spatially explicit wildfire hazard models, particularly in climate change hotspots like southwestern Europe, to support disaster risk reduction.
Key Innovation: Developing and evaluating a robust, globally transferable wildfire hazard assessment framework at a daily, 1-km scale by integrating biophysical and anthropogenic data (2007-2022) with machine learning models (ConvLSTM and XGBoost). The framework demonstrates strong predictive performance, identifies thermal and hydric forcings as key determinants, and offers a foundation for hybrid decision-support systems for resource allocation and early warning.
2. Correlation between macroscopic creep behavior and multi-scale pore characteristics of loess landslide soil: Experimental analysis and model prediction
Core Problem: Understanding the multi-scale pore evolution during the creep process of loess and its impact on shear strength, permeability, and ultimately, the stability of loess landslides, which is an important physical mechanism in these geohazards.
Key Innovation: Experimental analysis (creep and NMR tests) revealing the correlation between macroscopic creep behavior and multi-scale pore characteristics of loess landslide soil, identifying a positive-negative feedback mechanism affecting slope stability, and developing a more accurate macro-mesoscale creep model incorporating local bonding damage based on pore evolution.
3. A Human-Disaster Dynamic System Approach for Wildfire Adaptation
Core Problem: The debate on whether legislative measures effectively reduce housing damage in wildfires, particularly when existing models fail to consider socio-ecological feedback.
Key Innovation: Building a coupled land and hazard model to examine socio-ecological feedback to legislative measures, finding that structural/relief legislation can lead to increased development in wildfire-prone areas and greater damage, while educational/outreach legislation can offset these counterproductive effects by influencing risk perception and reducing community expansion in risk areas.
4. Study on shear strength prediction model of glass fiber-improved loess in seasonal frozen regions based on POA-XGBoost
Core Problem: Accurate prediction of shear strength for fiber-reinforced loess in seasonal frozen regions is challenging due to multiple interactive factors like freeze-thaw cycles, fiber content, and confining pressure.
Key Innovation: Developed a high-precision hybrid machine learning model (POA-XGBoost) for predicting the shear strength of glass fiber-improved loess, achieving superior predictive performance (R2 of 0.9767, MAPE of 0.0406) and significantly improving accuracy over benchmark models.
5. An impact-based drought classification method using real-world agricultural drought records and explainable automated machine learning
Core Problem: Accurate and practical drought assessment frameworks are critical for risk mitigation, but conventional methods may lack the ability to fully capture real-world impacts and causal drivers.
Key Innovation: Developed a novel impact-based drought classification framework synergizing causal inference (PCMCI+) with explainable AutoML (SHAP) using real-world agricultural drought records, outperforming conventional methods, identifying key climatic and non-climatic drivers, and demonstrating improved performance with antecedent climatic information.
6. A framework for mitigating boundary effects in transportation embankment seismic analysis: shaking table tests and numerical simulations
Core Problem: Boundary treatments in shaking table tests and finite element (FE) methods fundamentally influence the dynamic responses of transportation embankments during seismic analysis, leading to inaccuracies in predicting seismic behavior.
Key Innovation: Optimized the flexible boundary layer for scaled shaking table models and developed a viscoelastic boundary implementation for three-dimensional FE embankment models using EVA copolymer. The framework provides practical guidance for boundary design and highlights differences in dynamic responses between the two methods, improving seismic analysis accuracy.
7. A theoretical framework to describe undrained yielding and potential fluidisation of soft subgrade soil under cyclic loading
Core Problem: Soft soils subjected to undrained cyclic loading can experience either undrained yielding or fluidisation, leading to non-uniform changes in soil state, which needs a comprehensive theoretical framework to describe.
Key Innovation: Presents a theoretical framework combining empirical and critical state-based constitutive approaches to determine the evolution of excess pore water pressure, localized void ratio changes, and spatial state redistribution. It introduces a liquidity index criterion to identify potential localized fluidisation and uses critical state principles to assess undrained yielding, verified against experimental data.
8. Analysis of public opinions on natural disasters on short-video social media based on the Latent Dirichlet Allocation and prompt-BERT-TextCNN models
Core Problem: There is insufficient analysis of public opinion evolution regarding natural disasters disseminated through short-video social media platforms in the new media era.
Key Innovation: Proposed using Latent Dirichlet Allocation and prompt-BERT-TextCNN models to analyze public opinions on natural disasters on short-video social media, aiming to understand the evolution of public sentiment and information dissemination.
9. Multimodal remote sensing change detection: An image matching perspective
Core Problem: Multimodal Change Detection (MCD) struggles with designing per-pixel features truly invariant across modalities, making it difficult to accurately detect changes, especially in real-world disaster scenarios.
Key Innovation: IM4CD, a new unsupervised CD framework that unifies image matching and change detection by computing similarity via local template matching and utilizing spatial offset of response peaks to represent change intensity, integrating it with image co-registration, and leveraging modality-independent structural relationships via a Conditional Random Field (CRF) for dense change maps, showing robustness and potential for disaster response.
10. X-ray CT-based investigation of mesoscopic pore structure and macro-meso coupled damage model for anisotropic lean clay under a freeze–thaw cycle
Core Problem: The anisotropic mechanical properties of frozen soil, influenced by depositional conditions and skeleton particle arrangement, significantly affect the strength of frozen walls and the stability of subway tunnels constructed using artificial ground freezing (AGF) techniques, with current understanding lacking in quantifying freeze-thaw deterioration.
Key Innovation: Developed an apparatus to prepare anisotropic clay specimens with controlled sampling angles, used UCS tests and 3D CT scanning to investigate the influence of anisotropy and freezing temperature on mechanical properties and pore structure, quantified freeze-thaw cycle effects, and established a macro-meso coupled damage model to quantify soil deterioration for AGF-reinforced engineering applications.
11. Fully nonlinear analysis of wave diffraction by cylinders on a sloping terrain using the Irrotational Green-Naghdi model
Core Problem: Accurately modeling fully nonlinear wave diffraction and associated wave forces on structures (cylinders) located on sloping seabeds, and understanding the influence of various parameters.
Key Innovation: Development and validation of a finite element numerical model based on Irrotational Green-Naghdi (IGN) equations for fully nonlinear wave diffraction on sloping terrain, demonstrating the influence of seabed slope, cylinder spacing, and radius on wave forces and identifying wave resonance phenomena.
12. Accelerating the solution of Poisson equation in MPS method for tank sloshing using graph attention network
Core Problem: The high computational cost of solving the pressure Poisson equation in conventional Moving Particle Semi-implicit (MPS) simulations, especially for large-scale free-surface flow systems.
Key Innovation: Proposal of MPS-GAT, a data-driven hybrid approach integrating MPS with a Graph Attention Network (GAT) as a surrogate model for pressure estimation, significantly accelerating large-scale free-surface flow simulations (over 30x speedup) while maintaining accuracy and demonstrating strong generalization.
13. Numerical investigation of the secondary load cycle and high-frequency response on a vertical cylinder under focused waves
Core Problem: Understanding the characterization, mechanism, and impact of the secondary load cycle (SLC) induced by focused waves on the high-frequency dynamic response of vertical marine cylinders, which poses threats to structural safety.
Key Innovation: A 3D numerical wave tank study identifying the SLC as a strongly nonlinear, high-frequency load component primarily confined below the still water level, attributed to low-pressure vortex structures, and demonstrating its critical role in significantly amplifying high-frequency structural responses under extreme wave conditions.
14. Risk assessment of transboundary locust habitat distribution and migration pathways under climate change: a case study of Kazakhstan and Xinjiang, China
Core Problem: The transboundary habitat distribution and migration pathways of Calliptamus italicus and Locusta migratoria migratoria, and the associated risks under climate change, are not sufficiently understood, posing a threat to agricultural security along the China-Kazakhstan border.
Key Innovation: Conducted a risk assessment of transboundary locust habitat distribution and migration pathways under climate change, using a case study of Kazakhstan and Xinjiang, China, to address threats to agricultural security.
15. Governance and adaptive capacity: Greater losses in assets, flexibility, and confidence among tourism operators in low-governance contexts during the COVID-19 pandemic.
Core Problem: Understanding how adaptive capacity shifts in response to major systemic shocks, particularly across different governance contexts, to reduce disaster risk and strengthen resilience.
Key Innovation: First longitudinal quantification of multiple adaptive capacity domains (agency, assets, flexibility, learning, social organization, socio-cognitive constructs) across governance gradients during a non-environmental systemic shock (COVID-19). Findings reveal widespread asset declines but divergent patterns in human/psychological dimensions, with lower governance contexts experiencing greater losses in flexibility, agency, and confidence, highlighting governance as a key driver of adaptive trajectories.
16. Adaptive Thresholding (AT) for snowmelt detection with Calibrated Enhanced-Resolution Brightness Temperatures (CETB): Timing and regional patterns for case study of Alaska
Core Problem: Accurate and robust detection of snowmelt timing is critical for understanding hydrologic processes and monitoring environmental changes in heterogeneous snow-dominated regions, but legacy methods often lack accuracy across varying snow conditions.
Key Innovation: Introduces Adaptive Thresholding (AT), a data-driven framework that optimizes enhanced-resolution passive microwave brightness temperature (Tb) and diurnal amplitude variation (DAV) thresholds for snowmelt detection. AT achieved significantly higher accuracy (MAE ≤ 1.0 day) compared to legacy methods (MAE > 6-8 days) in interior Alaska, capturing spatial melt gradients more realistically and supporting large-scale monitoring.
17. Intelligent diagnostic framework for sustainable urban groundwater governance: Insights from asynchronous solute migration
Core Problem: Urban expansion and land transformation lead to decoupled and asynchronous solute migration in shallow aquifers, complicating groundwater quality management and making it difficult to identify and prioritize risk zones.
Key Innovation: Developed an integrated diagnostic framework combining process-based geochemical indicators (saturation indices, chloro-alkaline indices, Asynchronous Migration Index, Normalized Mutual Information) with interpretable, data-driven zoning (K-means, Gaussian mixture models, GeoSOM) to delineate hydrochemical regimes, identify early-warning zones, and diagnose groundwater-quality risks in rapidly urbanizing alluvial basins.
18. Hydrogeochemical and hydrogeological controls on spatiotemporal dynamics of groundwater phosphorus in a Lake Basin
Core Problem: The spatiotemporal dynamics and controlling mechanisms of phosphorus (P) contamination in groundwater within lake basins remain inadequately understood, despite its threats to human health and aquatic ecosystems.
Key Innovation: Employed an integrated approach (hydrochemical analysis, SOM, Eh-pH diagrams) to investigate groundwater total phosphorus (TP) dynamics in the Poyang Lake Basin. Revealed significant spatiotemporal heterogeneity, identified distinct regional controls (agricultural leaching, geogenic weathering, strongly reducing conditions promoting P release), and critically demonstrated that these processes are governed by the coupling of hydrogeological structures (e.g., confined aquifer, sluggish flow) and redox-sensitive geochemical processes, providing a basis for mechanism-based management.
19. A rapid field moisture measurement unit for compaction acceptance of unbound materials
Core Problem: Accurate and rapid field measurement of moisture content (MC) is critical for geomaterial compaction acceptance, but traditional methods like nuclear density gauges are becoming less desirable, and alternatives like lightweight deflectometers lack MC measurement.
Key Innovation: Developed and optimized a rapid field moisture measurement unit by improving an existing analyzer (Ohaus MB 120). This unit effectively mitigates field disturbances and accurately measures soil MCs, providing an effective means to evaluate compaction acceptance of unbound materials in situ.
20. A numerical study on effects of soil permeability and vibratory parameters on vibro-driving of open-ended piles in saturated sand
Core Problem: Uncertainties regarding the influence of soil and vibro-driving parameters on pile penetration rates and the evolving soil state during installation in saturated sand for offshore wind turbines.
Key Innovation: Systematic numerical investigation using large deformation modeling to understand the effects of soil permeability, dynamic force, vibration frequency, and hook load on pile penetration, displacement amplitudes, and changes in soil state (excess pore pressures, effective stresses) during vibro-driving.
21. Benchmark of plankton images classification: emphasizing features extraction over classifier complexity
Core Problem: The challenge of efficiently and accurately classifying vast datasets of plankton images, given their diversity, the prevalence of non-biological classes, and the rarity of many classes, compounded by a lack of systematic, realistic benchmarks for existing machine learning approaches.
Key Innovation: A comprehensive benchmark of plankton image classification approaches using large, realistic datasets, demonstrating that compact Convolutional Neural Networks (CNNs) are sufficient for extracting relevant features and that the quality of feature extraction is more critical than classifier complexity, providing insights for developing efficient operational classification models.
22. Guiding VI selection for phenology monitoring: Differential sensitivity of vegetation indices to temporal dynamics in canopy leaf area and pigment
Core Problem: Interpreting vegetation index (VI)-derived phenological metrics is challenging because canopy spectral signals simultaneously reflect both canopy structural changes (leaf area expansion) and leaf biochemical changes (leaf pigment dynamics).
Key Innovation: Integrated field measurements and PROSAIL model simulations to systematically assess how 21 VIs respond to canopy leaf area versus leaf pigmental changes, revealing distinct divergences in VI behavior and providing guidance for selecting appropriate VIs based on specific research goals.
23. Spatial occupancy index of tree crown: Can provide new perspectives for quantifying structural complexity of individual tree crowns
Core Problem: Existing methods for quantifying forest structural complexity lack incorporation of ecological niche occupancy and tree competitiveness, and often ignore in-depth analysis at the individual tree scale.
Key Innovation: Proposed a novel synergistic index, the spatial occupancy index of tree crown (SOI), and an optimized 3D alpha shape-voxelization (3D-ASV) algorithm for crown volume calculation. Demonstrated SOI's effectiveness in quantifying structural complexity and its high correlation with other canopy metrics, while also compensating for their limitations in explaining tree competitiveness.
24. Harnessing conditional generative adversarial networks for SAR-to-optical image translation via auxiliary geospatial landscape pattern-augmentation
Core Problem: SAR-to-optical image translation (S2OIT) is hindered by data heterogeneity and spectral discrepancies, and previous methods often generate low-fidelity images due to limited mining of geospatial landscape information.
Key Innovation: AGPA-CGAN, a conditional generative adversarial network framework with auxiliary geospatial landscape pattern-augmentation, which integrates SAR prior properties and geospatial structural information through an auxiliary pseudo-scattering pattern integration (APSPI) module and a geospatial landscape domain alignment (Geo-LDA) module, achieving high-quality S2OIT.
25. Hybrid process-based and deep learning for river nutrient prediction under limited monitoring data
Core Problem: Accurately simulating riverine nutrient dynamics (nitrogen and phosphorus) in catchments with limited monitoring data, where both standalone process-based models and data-driven approaches face limitations.
Key Innovation: Developed and evaluated hybrid modeling strategies combining PBM simulations (WSIMOD) with LSTM networks, demonstrating that these hybrid models outperformed empirical LSTM and standalone PBMs for river nutrient prediction. The study also provided insights into the value of extreme weather indices and different PBM outputs as input features, contributing to the interpretability of DL applications in water quality prediction.
26. Improving water quality prediction through landscape patterns in Euclidean distance-based drainage areas for plain river networks
Core Problem: Traditional spatial delineation methods are inadequate for assessing landscape-water quality relationships in complex plain river networks due to ambiguous flow directions and watershed boundaries, hindering accurate water quality prediction.
Key Innovation: Developed a Euclidean distance-based drainage area delineation method that assigns land cells to the nearest river segment, overcoming limitations of traditional methods. This approach, combined with machine learning (LightGBM), significantly improved water quality prediction accuracy (0.72) and identified spatially differentiated driver patterns and critical ecological thresholds for targeted management strategies.
27. Marine geotechnical investigation and soil testing for suction anchor foundation of floating wind turbines in the South China Sea
Core Problem: The increasing challenges in geotechnical investigations for deep-water floating wind turbine foundations, driven by project scope, time constraints, and demand for high-precision, reliable site data.
Key Innovation: Proposing a streamlined protocol for marine geotechnical investigations, emphasizing high-precision sampling at critical locations, and discussing consistency verification for geotechnical parameters in cyclic foundation design, offering practical guidance for optimizing procedures and developing durable foundations.
28. Biocementation of road pavements: An experimental investigation through physical modeling and accelerated testing
Core Problem: Deteriorating road pavements require non-disruptive rehabilitation solutions, and the effectiveness of biocementation for improving pavement structural performance needs direct proof of concept and investigation into its deformation mechanisms.
Key Innovation: Providing a direct proof of concept for biocementation in road pavements through full-depth physical modeling and accelerated testing, demonstrating significantly reduced deformation and enhanced structural integrity under higher loads, and revealing distinct subsurface deformation mechanisms via PIV analysis.
29. Activation of sintered red mud and its application as a cement substitute in foamed lightweight soil for backfilling engineering
Core Problem: Sintered red mud (SRM), a solid by-product, poses serious environmental challenges, and there is a need to enhance its pozzolanic potential to replace cement in foamed lightweight soil (FLS) for sustainable, large-scale backfilling applications.
Key Innovation: Developed a method to activate SRM, enabling its use as a cement substitute (up to 70% replacement) in FLS for backfilling. The developed SRM-FLS achieved increased unconfined compressive strength, good water stability, and deformation resistance, supported by a proposed mixed design framework for practical application.
30. Apparent Global Increase in Cloud Droplet Number Concentration After 2022 Attributed to MODIS Orbital Drift
Core Problem: The Aqua MODIS-retrieved cloud droplet number concentration (Nd) exhibits an unexpected and substantial increase over the near-global oceans after 2022, contradicting expected declines and distorting climate data records.
Key Innovation: This study demonstrates that the apparent surge in Nd is an artifact of sensor orbital drift and develops an empirical correction leveraging concurrent Suomi-NPP VIIRS observations to remove this artificial signal and quantify global mean Nd artificial biases.
31. Biases in Southern Ocean Precipitation From Shallow Convection: The Role of Cloud Morphology
Core Problem: Substantial discrepancies and uncertainties exist in satellite retrievals (GPM-IMERG, CloudSat) and reanalysis (ERA5) regarding precipitation from marine boundary layer clouds over the Southern Ocean, particularly in representing the contrasting precipitation between open and closed mesoscale cellular convection (MCC).
Key Innovation: Evaluated and quantified biases across datasets, showing GPM-IMERG underestimates precipitation from open MCCs and fails to reflect morphological distinction, while ERA5 captures the contrast but overestimates light precipitation from closed MCC. Highlighted the importance of improving the representation of marine atmospheric boundary layer cloud morphologies in models and observational datasets.
32. The Influence of Interactive Sea Surface Temperatures on Self‐Aggregation in Channel Domain Radiative‐Convective Equilibrium
Core Problem: Understanding how interactive sea surface temperatures (SST) influence convective self-aggregation in radiative-convective equilibrium (RCE) simulations, particularly the distinct aggregated states and feedback mechanisms.
Key Innovation: Conducted RCE simulations coupled to 2- and 20-m slab ocean models, finding that shallower slab oceans delay self-aggregation onset and lead to continuously evolving states, driven by low cloud-SST feedbacks and SST gradients promoting low-level moistening and convective initiation.
33. Dynamics of a Bio-Inspired Dual-Auger Burrowing Robot in Granular Media: DEM-MBD Cosimulation with Complementary Kinematics-Controlled Tests
Core Problem: Understanding the complex dynamics of bio-inspired burrowing robots in granular media and accurately simulating their interaction with the material.
Key Innovation: Development and validation of a DEM-MBD cosimulation approach to realistically model the fully coupled dynamics of a dual-auger burrowing robot in granular media, revealing insights into thrust, drag, and lift forces and the role of auger rotation.
34. Renewability of fossil groundwaters affected by present-day climate conditions
Core Problem: The common assumption that aquifer residence times directly indicate groundwater renewability is flawed, as the hydraulic response times (which control storage changes) and the influence of modern climates on fossil groundwater levels are often overlooked.
Key Innovation: Analysis of 31 major aquifers showed that water levels in many aquifers containing fossil groundwater are controlled by modern climates, demonstrating that evaluating renewability requires hydraulic analysis considering responses to abstraction and climate shifts.
35. Advancing airborne pollen mapping with integrated ground and satellite observations
Core Problem: Current understanding of atmospheric pollen is limited by sparse observational data, hindering accurate mapping of concentrations and composition across large regions.
Key Innovation: Developed a machine learning framework integrating spaceborne aerosol optical properties with meteorological and phenological data to estimate atmospheric pollen concentrations and composition across the continental U.S., demonstrating improved performance with satellite optical inputs and potential for data-sparse regions.
36. A novel scheme for estimating surface solar radiation based on Fengyun-2
Core Problem: Traditional surface solar radiation (SSR) estimation methods are limited by sparse ground stations, inconsistent data quality, and errors under complex cloud conditions, while physical retrievals are prone to errors.
Key Innovation: Proposed a DenseNet-based, cloud-transmittance-driven framework for SSR estimation using high-accuracy Himawari-8 radiation product as training reference, eliminating reliance on sparse ground stations and mitigating cloud-phase retrieval errors. Achieved superior performance compared to existing products.
37. Detecting global ocean subsurface density change with high-resolution via dual-task densely-former
Core Problem: The challenge of reconstructing high-resolution and high-reliability global ocean subsurface density to study dynamic processes and stratification under recent global ocean warming.
Key Innovation: Proposes DDFNet, a novel deep learning model using multi-scale feature extraction, attention mechanisms, and a dual-label design with an adaptive weighted loss function. It achieves high accuracy (R2 of 0.9863) and reveals a declining trend in global ocean subsurface density, facilitating studies on mesoscale ocean dynamics and climate change impacts.
38. Attributing GHG emissions to individual facilities using multi-temporal hyperspectral images: Methodology and applications
Core Problem: Current satellite remote sensing struggles to monitor GHG emissions from multiple densely clustered industrial sources effectively, hindering the implementation of mitigation policies.
Key Innovation: Proposes EA-MILES, an emission allocation framework integrating multi-source hyperspectral data with plume modeling to quantify process-level emissions from individual facilities. It shows good accuracy and detection coverage, providing transparent monitoring data to support mitigation and energy transition.
39. Beyond the surface: machine learning uncovers ENSO’s hidden and contrasting impacts on phytoplankton vertical structure
Core Problem: Satellite-based ocean remote sensing is limited to surface observations, preventing a comprehensive understanding of how the entire water column ecosystem responds to climate variability like ENSO, particularly regarding subsurface chlorophyll maximum (SCM) shifts.
Key Innovation: Develops and validates a novel stacked generalization ensemble machine learning framework to robustly reconstruct a 25-year high-resolution 3D chlorophyll-a field by integrating BGC-Argo profiles with multi-source satellite data. This framework reveals contrasting ENSO impacts on SCM and a significant subsurface-first response, demonstrating the necessity of a 3D perspective for climate variability assessment.
40. An automated approach to colorize soil maps
Core Problem: It is challenging to select distinguishable colors for complex soil maps that align with the hierarchical structure of soil taxonomy.
Key Innovation: Proposes an automated color selection algorithm for soil maps that encodes semantic relationships and adaptively generates colors through heuristic search, significantly outperforming expert-based methods in effectiveness and efficiency.
41. Time-series hydrochemical analysis of an interwell flow test through a fractured aquifer and comparison against microbial community data
Core Problem: Understanding the geochemical processes and flowpath heterogeneity during fluid injection/production in fractured rock masses to better manage geological reservoirs.
Key Innovation: Analyzed weekly time-series hydrochemistry data from an interwell injection test in a fractured metasedimentary rock formation, revealing distinct spatiotemporal patterns and evolving contributions from formation water and chemical reactions. It also demonstrated the complementary strengths of hydrochemical and microbial data for reservoir monitoring, with microbial data offering higher resolution as a "natural barcode".
42. Subsurface stormflow concentration-discharge relationships reveal DOC and nitrate transport mechanisms across land uses in karst hillslopes
Core Problem: The transport dynamics of dissolved organic carbon (DOC) and nitrate within the complex and hidden underground of karst hillslopes, especially across different land uses and during various rainfall events, remain unclear.
Key Innovation: Conducted high-frequency sampling to capture DOC/nitrate yields and their concentration-discharge relationships across subsurface and epikarst flow in karst hillslopes with four land-use types. Revealed that epikarst flow is the dominant pathway for nutrient export, identified land-use specific impacts on nutrient yields (e.g., forage grassland reducing exports), and elucidated mechanistic differences in DOC (source-limited) and nitrate (transport-limited with varying hysteresis) export, highlighting the primacy of epikarst systems and rainfall patterns for groundwater-quality protection.
43. A digital-twin LiDAR simulator for performance assessment of railway ballast geometry inspections
Core Problem: Testing and deploying autonomous inspection solutions based on LiDAR systems for railway infrastructure wear are challenging due to regulatory constraints and difficulty accessing representative railway environments.
Key Innovation: Developed a LiDAR digital twin simulator that realistically replicates sensor behavior in real railway track environments, incorporating CAD models and metrological calibration. This simulator enables performance assessment of ballast geometry measurement at different travel speeds, showing centimeter-order errors for commercial sensors up to 120 km/h.