Search Results (188)

Earthquake-induced soil liquefaction represents a critical geotechnical challenge due to its nonlinear soil–seismic interactions and its impact on structural safety. Traditional empirical methods often rely on simplified assumptions, limiting their predictive capability. This study develops and compares six machine learning (ML) classifiers—namely, Support Vector Machine (SVM), Artificial Neural Network (ANN), k-Nearest Neighbor (kNN), Random Forest (RF), Decision Tree (DT), and Naïve Bayes (NB)—to evaluate liquefaction susceptibility using an original dataset of 461 soil layers obtained from borehole penetration tests in the Edremit region (Balıkesir, NW Turkey). The models were trained and validated using normalized geotechnical and seismic parameters, and their performance was assessed based on accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve (AUC). Results demonstrate that SVM, ANN, and kNN consistently outperformed other models, achieving test accuracies above 93%, F1 scores exceeding 98%, and AUC values between 0.933 and 0.953. In contrast, DT and NB exhibited limited generalization (test accuracy of 84–88% and AUC of 0.78–0.82), while RF showed partial overfitting. In contrast, DT and NB exhibited weaker generalization, with test accuracies of 84% and 88% and AUC values of 0.78 and 0.82, respectively, while RF indicated partial overfitting. The findings confirm the superior capability of advanced ML models, particularly SVM, ANN, and kNN, in capturing complex nonlinear patterns in soil liquefaction. This study provides a robust framework and original dataset that enhance predictive reliability for seismic hazard assessment in earthquake-prone regions. Full article

Reclaimed coastal areas are highly susceptible to uneven subsidence caused by the consolidation of soft marine deposits, which can induce differential settlement, structural deterioration, and systemic risks to urban infrastructure. Further, engineering activities, such as construction and loadings, exacerbate subsidence, impacting infrastructure stability. Therefore, monitoring the integrity and vulnerability of linear urban infrastructure after construction on reclaimed land is critical for understanding settlement dynamics, ensuring safe and reliable operation and minimizing cascading hazards. Subsequently, in the present study, to monitor deformation of the linear infrastructure constructed over decades-old reclaimed land in Mokpo city, South Korea (where 70% of urban and port infrastructure is built on reclaimed land), we analyzed 79 Sentinel-1A SLC ascending-orbit datasets (2017–2023) using the Persistent Scatterer Interferometry (PSInSAR) technique to quantify vertical land motion (VLM). Results reveal settlement rates ranging from −12.36 to 4.44 mm/year, with an average of −1.50 mm/year across 1869 persistent scatterers located along major roads and railways. To interpret the underlying causes of this deformation, Casagrande plasticity analysis of subsurface materials revealed that deep marine clays beneath the reclaimed zones have low permeability and high compressibility, leading to slow pore-pressure dissipation and prolonged consolidation under sustained loading. This geotechnical behavior accounts for the persistent and spatially variable subsidence observed through PSInSAR. Spatial pattern analysis using Anselin Local Moran’s I further identified statistically significant clusters and outliers of VLM, delineating critical infrastructure segments where concentrated settlement poses heightened risks to transportation stability. A hyperbolic settlement model was also applied to anticipate nonlinear consolidation trends at vulnerable sites, predicting persistent subsidence through 2030. Proxy-based validation, integrating long-term groundwater variations, lithostratigraphy, effective shear-wave velocity (V_s³⁰), and geomorphological conditions, exhibited the reliability of the InSAR-derived deformation fields. The findings highlight that Mokpo’s decades-old reclamation fills remain geotechnically unstable, highlighting the urgent need for proactive monitoring, targeted soil improvement, structural reinforcement, and integrated InSAR-GNSS monitoring frameworks to ensure the structural integrity of road and railway infrastructure and to support sustainable urban development in reclaimed coastal cities worldwide. Full article

Rockburst disasters have become increasingly prevalent as distinct forms of subsurface geotechnical engineering advanced to the deep earth. Confronted with such a threatening subsurface geopressure disaster that poses a risk to personnel and equipment safety, the microseismic monitoring technology has been employed to track signals generated from rock fracture and collapse in the field. To guide the prevention and control of the hazard, the investigation conducted an effective microseismic data mining method. Through deep feature engineering and interpretable intelligence, a practical and available short-term prediction approach for the rockburst intensity class was developed. On the basis of rockburst case database collected from various underground geotechnical engineering, the neural network-based feature extraction method was conducted in the process of model training. The optimized model was obtained by combining the K-fold cross-validation approach with the structural parameter search methodology. The evaluation among the considered artificial intelligence models on the testing dataset was conducted and compared. Through analyses, the interpretable coupling intelligent model combining convolutional and recurrent neural networks for rockburst prediction were demonstrated with the most robust performance by evaluation metrics. Among them, the proposed adaptive feature extraction method leads the benchmark method by 6% for both accuracy and precision; meanwhile, the proposed metric generalization loss rate (GLR) for accuracy and precision in the validation–testing process reached 1.5% and 0.2%. Furthermore, the Shapley additive explanations (SHAP) approach was employed to verify the model interpretability by deciphering the model prediction from the perspective of the fined impact of input features. Therefore, the investigation demonstrates that the proposed method can predict rockburst intensity with robust generalization and feature extraction capabilities, which possess substantial engineering significance and academic worth. Full article

The proliferation of tunnel and subway networks in urban areas has heightened concerns regarding their vulnerability to seismic-induced liquefaction. This phenomenon, wherein saturated sandy soils lose strength and behave like a liquid under seismic waves, poses a catastrophic threat to the structural integrity and stability of underground constructions. While extensive research has been conducted to evaluate liquefaction triggering, most existing approaches rely on single ground motion intensity measures (e.g., PGA, I_A), which often fail to capture the combined effects of amplitude, energy, and duration on liquefaction behavior. In this study, the seismic response of saturated sandy soil from Yinchuan was analyzed using the Dafalias–Manzari constitutive model implemented in the OpenSeesPy platform. The model parameters were carefully calibrated using laboratory triaxial results. A total of ten real earthquake records were applied to evaluate two critical engineering demand parameters (EDPs): surface lateral displacement (SLD) and the maximum thickness of the liquefied layer (MTL). The results show that both SLD and MTL exhibit weak correlations with conventional intensity parameters, suggesting limited predictive value for engineering design. However, by applying Partial Least Squares (PLS) regression to combine multiple intensity measures, the prediction accuracy for SLD was significantly improved, with the correlation coefficient increasing to 0.81. In contrast, MTL remained poorly predicted due to its strong dependence on intrinsic soil characteristics such as permeability and fines content. These findings highlight the importance of integrating both seismic loading features and geotechnical soil properties in performance-based liquefaction hazard evaluation. Full article

In the North China region, measures such as restricting groundwater extraction and promoting cross-basin water diversion have effectively alleviated the problem of excessive groundwater exploitation. Nevertheless, the continuous rise in groundwater levels may alter the mechanical properties of foundation soil layers, potentially leading to geotechnical hazards such as foundation instability and the uneven settlement of structures. This study employs FLAC3D software to simulate the displacement, deformation, and stress–strain behavior of buildings and their surrounding strata during the dynamic recovery of groundwater levels, aiming to assess the impact of this process on structural integrity. Research findings indicate that the maximum building settlement within the study area reaches 54.8 mm, with a maximum inter-column differential settlement of 8.9 mm and a peak settlement rate of 0.16 mm/day. In regions where differential settlement aligns with the interface between the floor slab and walls, tensile stress concentrations are observed. The maximum tensile stress in these zones increases progressively from 1.8 MPa to 2.19 MPa, suggesting a potential risk of tensile cracking in the concrete structures. The influence of groundwater level recovery on buildings exhibits distinct phase characteristics, and the response mechanisms of different lithological strata vary significantly. Therefore, particular attention should be given to the physical properties and mechanical behavior of strata that are highly sensitive to variations in moisture content. These findings hold significant reference value for the sustainable development and utilization of underground space in the North China region. Full article

Geohazards pose significant dangers to human safety, infrastructures, and the environment, highlighting the need for advanced monitoring techniques for early damage detection and structure management. The distributed optical fiber sensors (DFOS) are strain, temperature, and vibration monitoring tools characterized by minimal intrusiveness, accuracy, ease of deployment, and the ability to perform measurements with high spatial resolution. Although these sensors rely on well-established measurement techniques, available for over 40 years, their diffusion within monitoring and early warning systems is still limited, and there is a certain mistrust towards them. In this regard, based on several case studies, the implementation of DFOS for early warning of various geotechnical hazards, such as landslides, earthquakes and subsidence, is discussed, providing a comparative analysis of the typical advantages and limitations of the different systems. The results show that real-time monitoring systems based on well-established distributed fiber-optic sensing techniques are now mature enough to enable reliable and long-term geotechnical applications, identifying a market segment that is only minimally saturated by using other monitoring techniques. More challenging remains the application of the technique for vibration detection that still requires improved interrogation technologies and standardized practices before it can be used in large-scale, real-time early warning systems. Full article

The objective of this study is to quantify and characterize ground deformations induced by underground coal mining in the Karaganda coal basin, Kazakhstan, in order to improve the understanding of subsidence processes and their long-term evolution. The SBAS-InSAR method was applied to Sentinel-1 (C-band) and TerraSAR-X (X-band) data from 2019–2021 to estimate the magnitude, extent, and temporal behavior of displacements over the Kostenko, Kuzembayev, Aktasskaya, and Saranskaya mines. The results reveal spatially coherent and progressive deformation, with maximum cumulative LOS displacements exceeding –800 mm in TerraSAR-X data within active longwall mining zones. Time-series analysis confirmed acceleration of displacement during active extraction and its subsequent attenuation after mining ceased. Comparative assessment demonstrated a strong agreement between Sentinel-1 and TerraSAR-X results (r = 0.9628), despite differences in resolution and acquisition geometry, highlighting the robustness of the SBAS-InSAR approach. Analysis of displacement over individual longwalls showed that several panels (3, 5, 8, 15, and 18) already exceeded their projected maximum subsidence values, underlining the necessity of continuous monitoring for ensuring safety. In contrast, other longwalls have not yet reached their maximum deformation, indicating potential for further activity. Overall, this study demonstrates the value of multi-sensor InSAR monitoring for reliable assessment of mining-induced subsidence and for supporting geotechnical risk management in post-industrial regions. Full article

Slope failures represent one of the most serious geotechnical hazards, which can have severe consequences for personnel, equipment, infrastructure, and other aspects of a mining operation. Deterministic and stochastic conventional methods of slope stability analysis are useful; however, some limitations in applicability may arise due to the inherent anisotropy of rock mass properties and rock mass interactions. In recent years, Machine Learning (ML) techniques have become powerful tools for improving prediction and risk assessment in slope stability analysis. This review provides a comprehensive overview of ML applications for analyzing slope stability and delves into the performance of each technique as well as the interrelationship between the geotechnical parameters of the rock mass. Supervised learning methods such as decision trees, support vector machines, random forests, gradient boosting, and neural networks have been applied by different authors to predict the safety factor and classify slopes. Unsupervised learning techniques such as clustering and Gaussian mixture models have also been applied to identify hidden patterns. The objective of this manuscript is to consolidate existing work by highlighting the advantages and limitations of different ML techniques, while identifying gaps that should be analyzed in future research. Full article

In recent years, controlling the integrity of shallow-buried natural gas pipelines within surface subsidence zones caused by high-intensity underground longwall mining in the Daniudi Gas Field of China’s Ordos Basin has emerged as a critical challenge impacting both mine planning and the safe, efficient co-exploitation of coal and deep natural gas resources. This study included field measurements and an analysis of surface subsidence data from high-intensity longwall mining operations at the Xiaobaodang No. 2 Coal Mine, revealing characteristic ground movement patterns under intensive extraction conditions. The subsidence basin was systematically divided into pipeline hazard zones using three key deformation indicators: horizontal strain, tilt, and curvature. Through ABAQUS-based 3D numerical modeling of coupled pipeline–coal seam mining systems, this research elucidated the spatiotemporal evolution of pipeline Von Mises stress under varying mining parameters, including working face advance rates, mining thicknesses, and pipeline orientation angles relative to the advance direction. The simulations further uncovered non-synchronous deformation behavior between the pipeline and its surrounding sand and soil, identifying two distinct evolutionary phases and three characteristic response patterns. Based on these findings, targeted pipeline integrity preservation measures were developed, with numerical validation demonstrating that maintaining advance rates below 10 m/d, restricting mining heights to under 2.5 m within the 260 m pre-mining influence zone, and where geotechnically feasible, the maximum stress of the pipeline laid perpendicular to the propulsion direction (90°) can be controlled below 480 MPa, and the separation amount between the pipe and the sand and soil can be controlled below 8.69 mm, which can effectively reduce the interference caused by mining. These results provide significant engineering guidance for optimizing longwall mining parameters while ensuring the structural integrity of shallow-buried pipelines in high-intensity extraction environments. Full article

Small landslides, though frequent, are often overlooked despite their significant potential impact on human-affected areas. This study presents an analysis of the Bella Orxeta landslide in Alicante, Spain, a rotational landslide event that occurred in March 2017 following intense and continued rainfall. Utilizing multitemporal datasets, including LiDAR from 2009 and 2016 and drone-based photogrammetry from 2021 and 2023, we generated high-resolution digital terrain models (DTMs) to assess morphological changes, estimate displaced volumes of approximately 3500 cubic meters, and monitor slope activity. Our analysis revealed substantial mass movement between 2016 and 2021, followed by relatively minor changes between 2021 and 2023, primarily related to fluvial erosion. This study demonstrates the effectiveness of UAV and DTM differencing techniques for landslide detection, volumetric analysis, and long-term monitoring in urbanized settings. Beyond its scientific contributions, the Bella Orxeta case offers pedagogical value across academic disciplines, supporting practical training in geomorphology, geotechnical assessment, GIS, and risk planning. It also highlights policy gaps in existing territorial risk plans, particularly regarding the integration of modern monitoring tools for small-scale but recurrent geohazards. Given climate change projections indicating more frequent high-intensity rainfall events in Mediterranean areas, the paper advocates for the systematic documentation of local landslide cases to improve hazard preparedness, urban resilience, and geoscience education. Full article

The Niland Moving Mud Spring, located near the southeastern margin of the Salton Sea, represents a rare and evolving geotechnical hazard. Unlike the typically stationary mud pots of the Salton Trough, this spring is a CO₂-driven mud spring that has migrated southwestward since 2016, at times exceeding 3 m per month, posing threats to critical infrastructure including rail lines, highways, and pipelines. Emergency mitigation efforts initiated in 2018, including decompression wells, containment berms, and route realignments, have since slowed and recently almost halted its movement and growth. This study integrates hydrochemical, temperature, stable isotope, and tritium data to propose a refined conceptual model of the Moving Mud Spring’s origin and migration. Temperature data from the Moving Mud Spring (26.5 °C to 28.3 °C) and elevated but non-geothermal total dissolved solids (~18,000 mg/L) suggest a shallow, thermally buffered groundwater source influenced by interaction with saline lacustrine sediments. Stable water isotope data follow an evaporative trajectory consistent with imported Colorado River water, while tritium concentrations (~5 TU) confirm a modern recharge source. These findings rule out deep geothermal or residual floodwater origins from the great “1906 flood”, and instead implicate more recent irrigation seepage or canal leakage as the primary water source. A key external forcing may be the 4.1 m drop in Salton Sea water level between 2003 and 2025, which has modified regional groundwater hydraulic head gradients. This recession likely enhanced lateral groundwater flow from the Moving Mud Spring area, potentially facilitating the migration of upwelling geothermal gases and contributing to spring movement. No faults or structural features reportedly align with the spring’s trajectory, and most major fault systems trend perpendicular to its movement. The hydrologically driven model proposed in this paper, linked to Salton Sea water level decline and correlated with the direction, rate, and timing of the spring’s migration, offers a new empirical explanation for the observed movement of the Niland Moving Mud Spring. Full article

Rainfall is a key triggering factor for numerous geotechnical hazards. Hence, it is necessary to investigate the degradation characteristics of rock–soil strength under different water contents. The existing macro–mesoscopic analysis methods for rock–soil strength degradation neglect the intrinsic connection between water content variations caused by external rainfall and mesoscopic mechanical mechanisms. In addition, there is a lack of discrete element method (DEM) mesoscopic parameter calibration methods for rock–soil strength under the influence of external environmental factors. Hence, this study aims to perform a macro–mesoscopic analysis and develop a parameter calibration model for the degradation of rock–soil strength under different water contents. First, the mesoscopic mechanical characteristics under different water contents are investigated by analyzing particle displacement, the bond failure rate, and the anisotropy coefficient. Interrelationships among shear strength, water content, and mesoscopic parameters are qualitatively analyzed, which indicated a macro–mesoscopic synergistic mechanism. A macro–meso-environment data set is constructed. Key mesoscopic parameters are determined using Pearson correlation (Pearson) and mutual information (MI) methods. Then, the mapping relationships are established based on ordinary least squares. The model accuracy is verified by comparing the calibrated simulation results with direct shear test results. The results show that the shear strength increases with vertical pressure under a constant water content. However, as the water content varies, the strength initially increases and then decreases. The average displacement of central particles and bond failure rate both decrease initially and then increase with rising water content, while the anisotropy coefficients show the opposite trend. Normal bond strength, tangential bond strength, and friction coefficient are determined as the key parameters. The goodness-of-fit R² of the parameter calibration model exceeds 0.92. Among 45 validation working conditions, only two are found to have errors of 12.4% and 13.6%, and the remainder have errors below 5%. Full article

Constructing a probabilistic assessment method is the primary task and key step in liquefaction research. This paper presents a systematic investigation into liquefaction potential evaluation methods. Through a comparative analysis of three conventional assessment methods, we identify critical limitations in existing approaches regarding accuracy and adaptability. A probabilistic ANN model was developed using field-collected standard penetration test (SPT) data from 311 liquefaction case histories. The model demonstrates superior performance with an overall accuracy of 86.17%, achieving 83.33% and 90.00% recognition rates for liquefied and non-liquefied cases, respectively. Key metrics, including precision (91.84%), recall (83.33%), and F1-score (87.38%), indicate robust discriminative capability. Comparative studies confirm the ANN model’s advantages over traditional methods in terms of prediction reliability and operational practicality. The research outcomes offer significant value for improving current liquefaction hazard assessment protocols in geotechnical engineering practice. Full article

Soil liquefaction poses a significant risk to foundations during earthquakes, prompting the need for simple, risk-aware screening tools that go beyond single deterministic boundaries. This study creates a probability-calibrated dual-threshold screening rule using a random forest (RF) classifier trained on 208 SPT case histories with quality-based weights (A/B/C = 1.0/0.70/0.40). The model is optimized with random search and calibrated through isotonic regression. Iso-probability contours from 1000 bootstrap samples produce paired thresholds for fines-corrected, overburden-normalized blow count N_1,60,CS and normalized cyclic stress ratio CSR_7.5,1 at target liquefaction probabilities P_liq = 5%, 20%, 50%, 80%, and 95%, with 90% confidence intervals. On an independent test set (n = 42), the calibrated model achieves AUC = 0.95, F₁ = 0.92, and a better Brier score than the uncalibrated RF. The screening rule classifies a site as susceptible when N_1,60,CS is at or below and CSR_7.5,1 is at or above the probability-specific thresholds. Designed for level ground, free field, and clean-to-silty sand sites, this tool maintains the familiarity of SPT-based charts while making risk assessment transparent and auditable for different facility importance levels. Sensitivity tests show its robustness to reasonable rescaling of quality weights. The framework offers transparent thresholds with uncertainty bands for routine preliminary assessments and to guide the need for more detailed, site-specific analyses. Full article

Geotechnical engineering faces challenges related to data, especially the ones related to dynamic soil behavior (i.e., shear modulus reduction and damping ratio curves with strain), with only a few datasets in open-access format and a slow transition to a more data-driven method. This lack of data, combined with variations in data collection methods, makes it difficult to build accurate predictive models. These challenges arose while developing a model to predict the shear modulus curves, an important soil property to better understand seismic hazard from three different databases. Combining multiple databases can sometimes degrade model performance. To address this, a novel approach in geotechnics based on Shapley values computed from an XGBoostRegressor model is introduced. This game–theoretic method quantifies each database’s marginal contribution to the model’s R² across all possible combinations, making it possible to identify which databases contribute most to improving performance. As the number of available databases continues to grow, this method will become increasingly useful. For shear modulus reduction curves, two out of three databases explored have Shapley values of 0.341 and 0.339, while the last one reaches only a value of 0.320. This suggests that the first two databases contribute more to the model’s performance. Full article