In large-scale underground caverns within jointed rock masses, the presence of multiple joints profoundly alters the overall strength and deformation characteristics of the rock, significantly impacting the stability of the surrounding rock. The discrete-based methods often suffer from low modelling and computational efficiency when applied to large-scale projects, while conventional continuum-based methods fail to accurately capture the governing influence of joints on the mechanical response of rock masses. To address these limitations, this study develops a three-dimensional Ubiquitous Multiple Joint Model (UMJM) that incorporates the mechanical behavior of multiple joint sets within a continuum framework. And the distance from the stress state point to the corresponding yield surface was used to determine the sequential yielding of multiple joint sets. Its accuracy and efficiency in strength prediction and deformation characterization is validated through uniaxial compression tests and circular tunnel excavation simulations. Subsequently, the UMJM is applied to the stability analysis of the left-bank underground caverns at the Baihetan Hydropower Station. The results demonstrate that the model accurately captures the large deformation behavior of surrounding rock under the interaction of joints and faults, with the computed deformations closely aligning with field monitoring data. Moreover, it reveals the controlling influence of joints on the evolution of plastic zones and stress perturbations. The UMJM enhances the realism of rock mass response simulations in jointed rock engineering without compromising computational efficiency, making it well-suited for large-scale underground engineering applications.

This study investigated the deformation behaviours, trigger mechanism and deformation pattern of Dawanzi (DWZ) slope, a bedding rock structure in the Baihetan reservoir area, based on the field geology surveys, site monitoring and geospatial statistical analysis methods. Monitoring results indicated that the displacement behaviours of measuring points TP01 and TP02 near the DWZ tunnel entrance area exhibited a creep deformation stage, while data from the other monitoring points remained stable, thereby confirming the effectiveness of the emergency anti-slide pile reinforcement. Furthermore, the correlations between slope stability and slope structure, gradient and lithology were analysed based on the statistical characteristics of 28 historical landslides in the Baihetan reservoir area. The statistical results revealed that the DWZ slope is a bedding rock slope with an inclination of approximately 35 degrees and a lithology characterised by a soft-hard interbedded sequence primarily composed of siltstone, silty mudstone and argillaceous dolomite, identifying it as a prone stratum. The rapid rise of the water level triggered bank collapse at the leading edge of the DWZ slope, which subsequently pulled the trailing edge rock mass forward along the slope, leading to continued deformation. The displacement normalisation method was applied to analyse the deformation pattern, indicating that the slope is currently in a steady-state deformation stage. This study provides a theoretical reference for the prevention and control of geological disasters in the reservoir area.

The high-speed movement characteristics of landslides represent a core scientific challenge in the fields of geotechnical engineering and geological hazards. Their essence lies in uncovering the physical mechanism underlying the reduction in shear strength during the sliding process. Existing thermal-hydro-mechanical (THM) coupling models often overlook the influence of hydrodynamic pressure on the dynamics of reservoir landslides, leading to discrepancies between simulation results and real-world scenarios. To address this gap, this study develops a theoretical framework for THM-coupled landslide dynamics that incorporates hydrodynamic pressure, grounded in the principles of mass conservation, energy conservation, and momentum. The framework employs the Crank-Nicholson finite difference scheme for discretizing the governing equations and utilizes the Thomas algorithm to solve the resulting tridiagonal matrix system, enabling dynamic simulation of temperature distribution in the shear zone, pore water pressure, and landslide kinematics. Furthermore, the effectiveness of the proposed model is validated through engineering case studies of the 2017 Xinmo landslide and the Yanshangou landslide in the Baihetan Reservoir, while the mechanism driving the high-speed movement of these landslides is analyzed in depth. Simulation results reveal two distinct softening mechanisms during landslide instability: frictional softening and THM coupling softening. These mechanisms exhibit phase-dependent dominance and synergistic effects throughout the sliding process. In conclusion, the proposed model provides reliable theoretical support for predicting the landslide dynamics in reservoir areas.

Landslide-generated impulse waves are among the most destructive natural hazards, especially in narrow rivers or enclosed reservoirs, often resulting in devastating consequences. This paper analyzes the deformation characteristics and triggering factors of the Tuandigou landslide through field investigations and on-site monitoring. The results show that landslide deformations are primarily manifested as ground cracks and concrete damage in the Dawanzi tunnel. Using the grey relation analysis method, impoundment is identified as the predominant factor influencing landslide deformations, followed by rainfall. By combining the granular flow model and renormalization group turbulence model in FLOW-3D, wave formation and propagation can be accurately described. After verification through grid resolution convergence tests, the numerical models are applied to predicting Tuandigou landslide-generated impulse waves. The results indicate that under the conditions of Maximum unstable landslide volume of 20× 106 m3 and highest water level of 825 m, the Maximum primary wave amplitude is 16.01 m, and the Maximum run-up on the riverbank opposite the landslide reaches 27.72 m. The wave reaching the dam front has a Maximum amplitude of 3.83 m, remaining below the dam crest, thereby ensuring structural safety of Baihetan Dam. Furthermore, a predictive formula for wave amplitude propagation is derived, revealing the attenuation law of the maximum wave amplitude along both upstream and downstream river sections. This research provides valuable insights into the risk of landslide-induced disasters in the Baihetan Reservoir area and other similar regions globally.

Since the impoundment of the Baihetan Reservoir, water-involved landslides have become widespread. Existing studies on landslide-generated waves have rarely examined the impact of varying water levels on wave characteristics. This paper focuses on the Wangjiashan (WJS) landslide in the Baihetan Reservoir area of China, conducting geomechanical experiments to investigate the spatiotemporal evolution of landslide-generated waves under different water level conditions. Utilizing a self-developed experimental measurement system, this study accurately records key parameters during the generation, propagation, and run-up of landslide-generated waves. It captures the complete sliding process of the WJS landslide under various water level conditions and elucidates the spatiotemporal distribution patterns of waves throughout their entire lifecycle, from generation through propagation to run-up. The research results indicate that water level factors significantly influence key parameters such as initial wave height, run-up on the opposite bank, propagation characteristics along the course, and maximum run-up in the Xiangbiling residential area. Generally, wave height initially increases and then decreases as the water level drops. Furthermore, this study offers crucial experimental data to deepen the understanding of the physical mechanisms of landslide-generated waves, advancing landslide disaster early warning technologies and enhancing the scientific accuracy and precision of landslide risk management.

In alpine and canyon areas, the problems of reservoir bank deformation and landslides are prominent. How to quickly and efficiently monitor deformation, timely identify geological disasters, and carry out surveys and treatments has become a crucial issue to be resolved urgently. Traditional survey methods are restricted by factors such as steep terrain and inconvenient transportation, resulting in low efficiency. In recent years, remote sensing technology has developed rapidly in the field of geological disaster monitoring, thanks to its high-precision deformation monitoring capabilities. This study is based on the Baihetan Hydropower Station during the water storage stage. The InSAR deformation monitoring technology is used to conduct large-scale disaster risk screening of the reservoir bank slopes. Firstly, the study analyzes the characteristics of UAV laser point clouds and images to construct a point cloud sequence. Subsequently, an improved iterative closest point (ICP) algorithm that integrates the scale-invariant feature transform (SIFT) and cylindrical neighborhood search is applied to improve the accuracy of slope deformation extraction. Finally, with the help of recognition algorithms and practical engineering experience, the surface deformation is analyzed based on the measured terrain data. The research shows that the surface deformation recognition technology based on UAV inspection and InSAR data has significant advantages in the monitoring of geological disasters in reservoirs in alpine and canyon areas. It can detect potential hazards in a timely manner and provide key decision-making basis for engineering projects.

This study mainly aimed to numerically investigate the influence of fractal roughness and pressure gradient on the fluid flow through a three-dimensional (3D) single rough fracture with a constant mechanical aperture. To avoid confusion created by the Reynolds number caused by different definitions of characteristic length, the critical pressure gradient was utilized to characterize the onset of non-linear flow in rough fractures. Further, the effect of fractal dimension on the critical pressure gradient was also investigated. The rough fracture surfaces with different fractal dimensions are generated by an open source code SynFrac, and the COMSOL Multiphysics software package based on the finite element method was used to simulate the fluid flow through the generated 3D rough fracture model by solving the Navier-Stokes equations. For each case, Forchheimer's law was used to describe nonlinear flow and the numerical results show that the critical pressure gradient, ranging from 15.68 kPa/m to 1.50 kPa/m, decreases exponentially with the increase in fractal dimension. The results also show that the relationship between relative effective fracture aperture and pressure gradient can be fitted well by a power function. These results suggest that the non-linear flow appears earlier as the fractal dimension and pressure gradient increase.

Introduction: Reservoir landslides undergo large deformations during the early stages of impoundment and maintain long-term persistent deformations during the operational period of the reservoir. The management of reservoir landslides mostly focuses on the early identification, risk assessment during the large deformations, and long-sequence monitoring during long-term persistent deformations, which requires sufficient continuity and integrity of the landslide monitoring data.Methods: Taking the Wulipo (WLP) landslide in Baihetan Reservoir as example, this paper proposes a reservoir landslide monitoring method that integrates field survey, unmanned aerial vehicle (UAV) photogrammetry and global navigation satellite system (GNSS) monitoring, which can effectively eliminate the practical monitoring gaps between multiple monitoring methods and improve the continuity and completeness of monitoring data.Results and discussion: First, this study determined the initiation time of the landslide through the field investigation and collected five period of UAV data to analyze the overall displacement vector of the WLP landslide using sub-pixel offset tracking (SPOT). On the basis of the above data, we compensated for the missing data in GNSS system due to the practical monitoring vacancies by combining the field survey and the landslide-water level relationship. Based on these monitoring data, this paper points out that the WLP landslide is a buoyancy-driven landslide, and whether or not accelerated deformation will occur is related to the maximum reservoir water level. Finally, this study analyzed and discussed the applicability of UAV photogrammetry for reservoir landslide monitoring in the absence of ground control points (GCPs), and concluded that this method can be quickly and flexibly applied to the stage of large deformation of reservoir landslides.

A hybrid smooth particle hydrodynamics (SPH) and shallow water equations (SWEs) model is proposed to simulate landslide-generated waves in river-valley reservoirs. The generation and propagation of impulse waves are simulated using the SPH model and SWEs model, respectively. A well-designed wave generation boundary is established as an interface between the SPH and the SWEs model. The Mohr–Coulomb constitutive model is used to represent the mechanical properties of the landslide. The accuracy of the proposed model are verified by the case of the Dayantang landslide. The proposed method is then applied to predict impulse waves generated by the Wangjiashan landslide. The results show that at the normal water level of 825 m, the maximum sliding velocity is 7.30 m/s, the height of the leading wave is about 7.93 m, and the runup height at the opposite bank is about 5.32 m. The impulse wave reaches the Xiangbiling settlement with a maximum runup height about of 4.54 m, which is higher than the maximum design elevation of 827.50 m of the settlement. It is recommended that disaster risk reduction measures should be implemented in the Wangjiashan landslide. The proposed hybrid SPH-SWEs model provides an effective tool for predicting landslide-generated waves.

The risk management of landslide surges after water storage in large reservoirs is a major challenge in reservoir management. The water storage in the reservoir area of the Baihetan hydropower station on the Jinsha River brings the potential of Wangjiashan landslide activation. This paper proposes a mitigation measurement based on slope cutting for the Wangjiashan landslide and evaluates the feasibility and effectiveness of the management scheme through numerical simulation. Results indicate that when water storage is in a normal level, after an earthquake, the landslide will impact the Baihetan Reservoir with a maximum velocity of 8.52 m/s, generating waves with a maximum wave height of 15.63 m. By removing the upper part of the landslide, the volume of the landslide is reduced from 611x104m3 to 369x104m3, it will cut down the maximum wave height in the Xiangbiling residential area to 3.09 m, which significantly reduces the risk of landslide surge waves. It further establishes a disaster prevention and control plan for the landslide-induced wave disaster. Overall, this study provides important theoretical and practical references for the risk management of landslide surge waves and offers valuable insights for addressing similar issues in the future.

The Baihetan Hydropower Station reservoir area began impoundment in 2021, triggering the reactivation of ancient landslides and the formation of new ones. This not only caused direct landslide disasters but also significantly increased the likelihood of secondary surge wave disasters. This study takes the Wangjiashan (WJS) landslide in the Baihetan reservoir area as an example and conducts large-scale three-dimensional physical model experiments. Based on the results of the physical model experiments, numerical simulation is used as a comparative verification tool. The results show that the numerical simulation method effectively reproduces the formation and propagation process of the WJS landslide-induced surge waves observed in the physical experiments. At the impoundment water level of 825 m, the surge waves generated by the WJS landslide pose potential threats to the Xiangbiling (XBL) residential area. In this study, the numerical simulation based on computational fluid dynamics confirmed the actual propagation forms of the surge waves, aligning well with the results of the physical experiments at a microscopic scale. However, at a macroscopic scale, there is some discrepancy between the numerical simulation results and the physical experiment outcomes, with a maximum error of 25%, primarily stemming from the three-dimensional numerical source model. This study emphasizes the critical role of physical model experiments in understanding and mitigating surge wave disasters in China. Furthermore, physical experiments remain crucial for accurate disaster prediction and mitigation strategies. The theories and methods used in this study will provide important references for future research related to landslide disasters in reservoir areas.

The large reservoirs in the southwestern Alpine Canyon region are characterized by long reservoir banks and complex geological structures. The problem of finding the deformation zone quickly and efficiently is urgent and needs to be resolved. In this study, taking the area 110 km upstream of the Baihetan dam site as the study area, the applicability of various interferometric synthetic aperture radar (InSAR) techniques was summarized, and the small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) method was used to carry out large-scale disaster risk screening using multi-source satellite SAR data. A total of 40 hidden danger points were identified, with 22 of them being newly discovered. The differences in the deformation results from the multi-source satellite SAR data were discussed. By analyzing ComparSAR-based deformation monitoring results of a large reservoir, a new risk screening method for slopes in alpine-canyon regions can be provided.

The deformation and failure of the surrounding rock mass is a key issue during the construction of large‐scale underground powerhouse, and large discontinuities are likely to cause this problem in the presence of complex geological structures. This article takes the right bank underground powerhouse of the Baihetan Hydropower Station as a case study. In this case, deformation mutation of the surrounding rock mass occurred in the south section of the main powerhouse, with the maximum deformation reaching 178 mm, and the deformation and failure of different parts showed differences. A comprehensive study integrating field survey, site monitoring, laboratory test and numerical simulation was carried out. By field survey and monitoring, characteristics of deformation and failure are described, and the spatiotemporal difference in deformation is analysed. The stress evolution during excavation is studied based on numerical simulations, the mechanical response of rock is derived through laboratory tests, and the mechanism of spatiotemporal difference is revealed. The results indicate that the main reason for the spatiotemporal difference is the presence of slightly inclining interlayer shear zone C 4 . In the south section, the excavation‐induced stress concentration at the arch was enhanced due to C 4 , with the maximum principal stress exceeding 70 MPa, and the high compressive stress here triggered the deformation mutation of surrounding rock mass. After undergoing a stress path from concentration to unloading, the surrounding rock mass at the downstream sidewall was seriously damaged, and its deformation also mutated under approximately vertical stress. The mutation resulted in the uneven spatial distribution, large increment and time‐dependent feature of deformation.

Landslide-induced waves pose significant risks to human lives, property, and infrastructure. The multifaceted nature of landslide movements combined with solid-fluid interactions makes hazard assessment of these waves particularly challenging. This study proposes a novel hybrid numerical method for simulating potential landslide-induced wave. The Material Point Method (MPM) is employed to evaluate landslide movement, taking into account residual slope strength and examining a potential landslide's displacement and velocity. Concurrently, the Smoothed Particle Hydrodynamics (SPH) method is used to model the propagation characteristics of impulse waves. This novel method's validity is established through two physical tests. Subsequently, it is employed to assess the impulse wave risk arising from a potential landslide in China. The findings indicate a maximum wave amplitude of 5.661 m, which reaches a proximate residential zone 75 s post-landslide, traveling at an approximate speed of 2.5 m/s. The influence of a landslide's residual strength on the amplitude of an impulse wave is also explored. Notably, a 20% increase in residual strength results in a 62% reduction in peak wave amplitude. Hence, comprehensive geotechnical investigations and tests are indispensable for gauging the risks associated with landslide-induced waves. This research offers a potent numerical simulation technique for such waves and furnishes valuable insights into risk assessment.

When excavating the rock foundation of a hydropower station, it will be affected by the phenomenon of unloading and relaxation, which may increase the risk of stability of the dam foundation engineering system. The dam foundation of Baihetan Hydropower is a columnar jointed rock mass (CJRM), which presents strong brittleness and anisotropy compared to traditional dam foundation rocks. Therefore, this type of rock mass is prone to disturbance to the dam body, structure, etc. during excavation, so it is necessary to accurately evaluate the impact of dam foundation excavation. Establishing a rock mass creep models serve as an effective tool for evaluating such stability but often suffer from significant parameter uncertainty. Digital twin technology, a virtual model, is capable of real-time learning from actual monitoring data obtained from the physical entity to enhance the performance of the built-in mechanistic model. In this study, the researchers employ the classical Burgers constitutive equation as the theoretical framework and integrate it with an ensemble smoother with multiple data assimilation (ESMDA) method based on Bayesian principles, along with displacement monitoring data from the Baihetan Dam foundation, to construct a digital twin model. Within this framework, the researchers analyze the uncertainty of rheological parameters at various measurement points in the Baihetan Dam foundation. Subsequently, the most suitable rheological parameters are selected and incorporated into the constitutive model to obtain displacement estimates, which are then compared with on-site monitoring data. The results demonstrate that the proposed method effectively performs probabilistic parameter estimation and model prediction for rheological mechanics. This research integrates data-driven methods with mechanical principles, offering a reliable approach for assessing the uncertainty of unloading rheological parameters and displacement prediction in dam foundations, thereby providing essential support for the evaluation of excavation projects in the CJRM of the Baihetan Dam foundation.

In evaluating the safety of rock slopes engineering, it is imperative to account for rheological effects. These effects can lead to significant deformations that may adversely impact the overall structural integrity. Consequently, accurate determination of the rheological mechanical parameters of slope rocks is essential. However, the application of rheological parameters obtained from laboratory tests encounters limitations due to the rock's inherent heterogeneity, scale effects, and inevitable sample dispersion. By contrast, on-site monitoring data serve as critical assets for real-time calibration and risk assessment in the evaluation of rheological parameters and prediction of slope deformation. To integrate on-site monitoring data with rheological mechanical mechanisms, this study introduces a probabilistic inverse model for evaluating rock slope rheological parameters, grounded in Bayesian theory, and incorporating a No-U-Turn Sampler (NUTS) based on Markov Chain Monte Carlo (MCMC) sampling algorithm. In terms of methodological efficiency, we compared the NUTS method with the traditional Metropolis-Hastings (M-H) approach, demonstrating the superior efficiency of the former. Additionally, sensitivity analysis of rheological parameters was conducted using the Burgers constitutive model. By combining the NUTS-based MCMC method with this model, the uncertainty of creep parameters was successfully evaluated. Utilizing these updated posterior parameters, up to 3-year deformation forecast for the slope was executed, the findings demonstrate that the deformation on the left bank slope is slight, indicating a state of safety. This study integrates monitoring data with rheological mechanics to establish a physical-data-driven rheological safety assessment mechanism. It offers a scientifically robust and effective approach for the uncertainty evaluation of rheological parameters and deformation prediction, providing significant support for the safety assessment of the left bank slope of the Baihetan hydropower station, China.

The task of landslide recognition focuses on extracting the location and extent of landslides over large areas, providing ample data support for subsequent landslide research. This study explores the use of UAV and deep learning technologies to achieve robust landslide recognition in a more rational, simpler, and faster manner. Specifically, the widely successful DeepLabV3+ model was used as a blueprint and a dual-encoder design was introduced to reconstruct a novel semantic segmentation model consisting of Encoder1, Encoder2, Mixer and Decoder modules. This model, named DeepLab for Landslide (DeepLab4LS), considers topographic information as a supplement to DeepLabV3+, and is expected to improve the efficiency of landslide recognition by extracting shape information from relative elevation, slope, and hillshade. Additionally, a novel loss function term—Positive Enhanced loss (PE loss)—was incorporated into the training of DeepLab4LS, significantly enhancing its ability to understand positive samples. DeepLab4LS was then applied to a UAV dataset of Baihetan reservoir, where comparative tests demonstrated its high performance in landslide recognition tasks. We found that DeepLab4LS has a stronger inference capability for landslides with less distinct boundary information, and delineates landslide boundaries more precisely. More specifically, in terms of evaluation metrics, DeepLab4LS achieved a mean intersection over union (mIoU) of 76.0% on the validation set, which is a substantial 5.5 percentage point improvement over DeepLabV3+. Moreover, the study also validated the rationale behind the dual-encoder design and the introduction of PE loss through ablation experiments. Overall, this research presents a robust semantic segmentation model for landslide recognition that considers both optical and topographic semantics of landslides, emulating the recognition pathways of human experts, and is highly suitable for landslide recognition based on UAV datasets.

An anisotropic rheological damage constitutive model for columnar jointed rock mass (CJRM) is proposed based on the microstructure tensor theory. Based on the variation of rheological strain curve over time, the relationship between damage and rheological strain is established by considering the influence of time effect on the deformation of CJRM, and the aging damage variable is introduced. The second development of the self-defined constitutive model based on FlAC3D is established, and the validity of the model is verified by comparing the experimental results with the numerical results. Furthermore, the developed model is applied to Baihetan engineering slope analysis and the results show that the anisotropic characteristics of rock mass containing columnar joints are closer to field monitoring results during the long-term deformation process rather than the isotropic deformation characteristics. The results obtained serve as an insightful reference point for Baihetan engineering slope analysis.

Landslide disasters pose a significant threat, with their highly destructive nature underscoring the critical importance of timely and accurate recognition for effective early warning systems and emergency response efforts. In recent years, substantial advancements have been made in the realm of landslide recognition (LR) based on remote sensing data, leveraging deep learning techniques. However, the intricate and varied environments in which landslides occur often present challenges in detecting subtle changes, especially when relying solely on optical remote sensing images. InSAR (Interferometric Synthetic Aperture Radar) technology emerges as a valuable tool for LR, providing more detailed ground deformation data and enhancing the theoretical foundation. To harness the slow deformation characteristics of landslides, we developed the FCADenseNet model. This model is designed to learn features and patterns within ground deformation data, with a specific focus on improving LR. A noteworthy aspect of our model is the integration of an attention mechanism, which considers various monitoring factors. This holistic approach enables the comprehensive detection of landslide disasters across entire watersheds, providing valuable information on landslide hazards. Our experimental results demonstrate the effectiveness of the FCADenseNet model, with an F1-score of 0.7611, which is 9.53% higher than that of FC_DenseNet. This study substantiates the feasibility and efficacy of combining InSAR with deep learning methods for LR. The insights gained from this research contribute to the advancement of regional landslide geological hazard monitoring, identification, and prevention strategies.

A retrogressive landslide is influenced by the cyclical fluctuations in reservoir water levels is considered a common natural disaster. Tension cracks are important indicators for assessing landslide status in the case of retrogressive landslides. Displacement monitoring is a commonly used method and provides an intuitive reflection of the landslide deformation; however, it does not directly indicate the depth of the tension cracks. Based on the principles of vibrational dynamics, a retrogressive landslide is proposed to be initially classified as a single-mass spring oscillator model before the development of cracks. Following the development of tension cracks, the model can be classified as a double-mass spring oscillator model. The model patterns are verified through numerical simulations using ABAQUS. Based on the numerical simulations, with an increase in the number of reservoir water cycle fluctuations, the displacement and stress of the landslide exhibit periodic growth. However, during displacement growth, the tension cracks do not necessarily increase. As the tension cracks deepen, the landslide transitions from a single-mass spring oscillator model to a double-mass spring oscillator model, with the appearance of a second-order natural frequency. Moreover, as the tension cracks deepen, the numerical values of the natural frequency change. The maximum change in first-order natural frequency is 3.5 Hz. The maximum change in second-order natural frequency is 4.5 Hz. The variation in the natural frequency can reflect the depth of development of the landslide's tension cracks and, consequently, indicate changes in the stability state of the landslide.