Next Article in Journal
Assessing Marine Resource Carrying Capacity: Methods, Economic Impacts, and Management Strategies
Previous Article in Journal
An Innovative Device for the Hot Water Circuit in Hospitals to Save Energy Without Compromising the Safety and Quality of Water: Preliminary Results
Previous Article in Special Issue
Mechanical Properties and Energy Evolution Laws of Rocks Under Freeze–Thaw
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 17
Issue 5
10.3390/w17050695
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Numerical Investigation into the Runout Dynamics of Reservoir Landslides: Insights from the Yanguan Landslide

by
Hao Fang
1,2,3,
Bing Li
4,*,
Kai Liu
2 and
Yaobin Meng
2
1
Zhejiang Institute of Geosciences, Hangzhou 310007, China
2
School of National Safety and Emergency Management, Beijing Normal University, Beijing 100875, China
3
Qiantang Institute of Geosciences, Hangzhou 310007, China
4
China Institute of Geo-Environment Monitoring, Beijing 100081, China
*
Author to whom correspondence should be addressed.
Water 2025, 17(5), 695; https://doi.org/10.3390/w17050695
Submission received: 24 January 2025 / Revised: 22 February 2025 / Accepted: 23 February 2025 / Published: 27 February 2025
(This article belongs to the Special Issue Approaches to Water-induced Landslide Hazard Risk Forecasting and Assessment)
Browse Figures
Versions Notes

Abstract

:
Understanding the dynamic behavior of landslides is essential for effective risk assessment. This study examines the Yanguan landslide, which occurred on 29 October 2017, in the Three Gorges Reservoir (TGR) region of China. Due to its unique capability in modeling discontinuum behaviors during landslide fragmentation, the discrete element method was utilized to analyze the movement characteristics of this landslide. The investigation began with a field survey to assess the geological features and failure mechanism of the landslide, which indicates that the landslide was likely triggered by prolonged variations in reservoir water levels and heavy rainfall preceding the event. Following this, a three-dimensional numerical model of the landslide was constructed using pre- and post-event terrain data. The accuracy of the numerical model was validated by comparing its simulation results with field survey data. Finally, the landslide’s movement behavior and energy transformation were analyzed based on the validated model. This work can enhance landslide risk assessment by quantifying dynamic parameters critical for impact prediction, further provide a scientific basis for the study of the landslides in the TGR area, and contribute to disaster prevention.

1. Introduction

Landslides constitute one of the most detrimental geohazards globally, inflicting catastrophic damage on infrastructure and resulting in significant loss of human life [1,2]. Global landslide fatalities exceeded 32,000 annually (2010–2022 average) according to WHO disaster databases, with reservoir-related slope failures accounting for 18% of these incidents in Asia [3,4]. Over the past several decades, numerous landslides with severe repercussions have transpired in the Three Gorges Reservoir (TGR) area of China [5,6]. Consequently, extensive research has been conducted on the failure mechanisms [7], monitoring and early-warning techniques [8,9,10], triggering factors [11], and mitigation strategies of these landslides [12]. However, investigations into the runout behavior of landslides in this region remain comparatively scarce [13,14].
The runout process of a landslide is an intricate mechanical phenomenon. A variety of methods have been developed to analyze this behavior, encompassing field surveys [14,15], empirical and/or analytical solutions [16,17], experimental tests [18,19], and numerical simulations [20,21]. Field surveys, while informative, often necessitate substantial financial investment and may yield only qualitative outcomes [22]. For example, conventional field investigations typically involve geotechnical professionals conducting on-site surveys through geophysical surveys and borehole drilling to acquire geomechanical parameters of landslide masses [23]. However, this methodology has inherent limitations in reconstructing the complete dynamic mechanisms of slope instability, particularly in capturing the progressive failure stages and real-time deformation patterns [24]. Empirical and analytical solutions tend to rely on numerous assumptions and may apply only to specific situations [25]. For complex landslides that cannot be accurately assessed using empirical or analytical methods, experimental tests provide an alternative means of analyzing their runout behavior [26], albeit with potential concerns regarding cost efficiency [27]. In light of these considerations, numerical simulations emerge as a viable compromise for analyzing landslide runout behavior [28]. Numerical simulations achieve an optimal accuracy-feasibility balance by resolving complex geological interactions at scales impractical for physical experiments while avoiding the oversimplifications inherent in purely analytical approaches, making them uniquely suited for reconstructing multi-stage landslide dynamics under real-world constraints, as evidenced by studies on the Liangshuicun landslide in China [29], the Oso landslide in the USA [30], and the long-runout disaster chain triggered by rock and ice avalanche at Chamoli, India [31].
In fact, contemporary landslide studies often employ fragmented methodologies, predominantly focusing on isolated mechanisms while neglecting integrated approaches to address cascading failure processes [32]. Regarding numerical modeling of reservoir landslide dynamics, prevalent limitations persist in achieving sufficient accuracy, particularly in resolving granular-scale interactions, alongside computational inefficiencies that hinder high-fidelity replication of multistage deformation processes.
In recent years, numerical simulation methods have emerged as a powerful tool for analyzing landslide runout behavior, providing critical insights into propagation dynamics, predicting runout distances, and conducting comprehensive risk assessments [33,34,35]. Among these methods, the Particle Flow Code (PFC) has gained widespread recognition for its ability to accurately simulate landslide runout processes [36,37,38]. As a discrete element method (DEM)-based numerical tool, PFC excels in landslide simulation by modeling granular particle interactions, effectively capturing large deformations, which can reconstruct the complete landslide process from fracture initiation to debris deposition. This paper presents a case study using PFC to simulate the runout behavior of the Yanguan landslide in the Three Gorges Reservoir (TGR) region of China. The Yanguan landslide, which occurred on 29 October 2017, in Guizhou Town, Zigui County, resulted in an economic loss of approximately RMB 4.5 million [39]. This study integrates field survey data and UAV photogrammetry to examine the failure mechanism of the landslide and construct a three-dimensional numerical model. The contact micro-parameters within the PFC model are calibrated using a combination of support vector machine (SVM) analysis and uniaxial compression simulations. The study then numerically investigates the runout behavior of the Yanguan landslide and explores its deposition characteristics. The primary contribution of this research is the detailed analysis of the failure mechanism of a reservoir-induced landslide in the TGR region and the simulation of its runout behavior.
This study advances reservoir landslide research by establishing an integrated geomechanical framework that systematically resolves terrain–material interactions governing multistage failure dynamics through coupled field monitoring and enhanced discrete element modeling. The work specifically aims to (1) elucidate progressive deformation mechanisms via spatiotemporal velocity-energy coupling analysis, and (2) develop a high-resolution numerical protocol capable of capturing granular-scale interaction effects during hydrodynamic loading processes, thereby addressing critical gaps in mechanistic understanding and simulation fidelity. In this paper, we first introduce the geological background and failure mechanism of the Yanguan landslide. Subsequently, the methodology to construct the 3-D landslide model is described in detail, including the calibration of the model parameters. Next, the simulation results from the PFC model are compared with field survey data to verify the model’s accuracy. Following this, the landslide mobility dynamics and energetic evolution mechanisms are investigated through the calibrated computational framework, culminating in conclusions derived from validated numerical evidence.

2. Study Area

Within this segment, an overview is provided regarding the geological context and the collapse dynamics of the Yanguan landslide. Leveraging the outcomes from field inspections and images captured via unmanned aerial vehicle (UAV), the geological circumstances of the research locale and the distinct features of the Yanguan landslide are elucidated. Subsequently, an exploration into the failure process of this landslide is conducted, drawing upon field investigation findings, thereby facilitating an in-depth analysis of its failure mechanism.

2.1. Overview of the Research Area

The Yanguan landslide occurred in Zigui County, Yichang City, China, approximately 12 km downstream from the mouth of the Xiangxi River (refer to Figure 1a). This region is characterized by mid-to-high mountains and precipitous valleys. The peak elevation in this area reaches around 1200 m above sea level (asl), with relative height differences ranging from 280 to 960 m. The predominant geological strata in the study area are the Middle Jurassic clastic sedimentary rocks, commonly referred to as the red layer [40], and the typical slope gradient lies between 20 and 40°. There are no significant faults in the vicinity (see Figure 1b), and the sedimentary strata exhibit stability. The exposed strata primarily consist of the Middle Jurassic Niejiashan formations (J2n), which comprise interbedded purple-red, thick-bedded siltstone and fuchsia medium-to-thin silty mudstone. The Yanguan slope features oblique bedding, oriented at 115°. Sandstone outcrops form ridges, whereas mudstone outcrops are relatively gentler (see Figure 2). The Qel distribution predominantly occurs in mudstone outcrop areas, creating very gentle slopes. Apart from exposed sandstone cliffs and road cuts, the slope vegetation is well developed, primarily consisting of trees and citrus groves. To sum up, the sedimentary environment in the zone is stable, and the tectonic activity in this area likely has minimal impact on the instability of this landslide.

2.2. Geological Conditions

The post-failure topography of the Yanguan landslide is depicted in Figure 2. This “tongue-shaped” landslide, situated between elevations of 156 m to 328 m, has its front edge bordered by the Xiangxi River and its rear edge by a steep bedrock mountain. The landslide deposits elevations span from 150 m to 280 m, averaging a thickness of 18 m. The landslide encompasses an area of 69,191 m2, with a volume of 1.25 × 104 m3, stretching approximately 374 m in length and 175 m in width [39].
The sliding mass features three-tiered platforms extending from the front to the rear, as shown in Figure 3. In the longitudinal section, the upper part exhibits a zigzag pattern, while the lower part forms an arc, with the mass being thinner at the ends and thicker in the middle, reaching a maximum thickness of 24.7 m. In the horizontal section, both parts show an arc trend, thin at the sides and thickening towards the center. The mass is identified as Holocene deposits, primarily consisting of brownish-red to gray-yellow silty clay and mudstone gravel blocks.
Field investigations reveal that the slip zone soil is purple-red gravelly loam and clay, predominantly made up of purple-red siltstone and gray-green calcareous sandstone. The slide bed, where the sliding mass deforms and slides, is the underlying bedrock layer. The rock formation has a strike of approximately 295° and a dip of 31°, with clearly visible bedding planes in outcrops.

2.3. Meteorological and Hydrological Settings

The study area experiences pronounced seasonal variations in precipitation [41], as depicted in Figure 4. During the dry season from December to February, rainfall is minimal, averaging approximately 40 mm per month, with the lowest monthly total recorded at just 4 mm. Following the conclusion of the rainy season (September to October), precipitation decreases, although recent years have seen an uptick, with cumulative rainfall in 2017 reaching 395.6 mm. The Xiangxi River traverses the leading edge of the landslide area. Its water level is synchronized with that of the Yangtze River and fluctuates periodically due to the influence of the Three Gorges Reservoir (TGR). Before the onset of the rainy season, the TGR water level drops from 175 m to 145 m. After the rainy season concludes, the TGR water level rises from 145 m to 175 m. The elevated reservoir water level following the rainy season, coupled with heavy rainfall, creates a “special hydrological environment” that promotes increased deformation of existing landslides [39], such as the Shuping landslide [42], and facilitates the occurrence of new landslides, including the Yanguan landslide.

2.4. Failure Process and Mechanism

After the failure of the Yanguan landslide, a UAV survey was conducted, and the resulting motion characteristic analysis map is presented in Figure 5. This figure primarily illustrates the extent, direction, and distance of the landslide. As depicted in Figure 5, the Yanguan landslide experienced two consecutive slips. The initial slip occurred predominantly near the S255 County Road Bridge, while the second slip covered a broader area. The first slip, traveling approximately 70 m in an N82°W direction, damaged a 50 m stretch of road and distorted a building. The second slip, moving in an N75°W direction, inflicted significant damage on the concrete pavement. Field observations revealed that the damaged road was piled up at the forefront (see Figure 5), yet remnants of the S255 County Road Bridge were found submerged in the river, suggesting that the bridge debris may have traveled further than the damaged road. This implies that the deposits from the first slip were likely propelled by the momentum of the second slip. Note that the second sliding of the Yanguan landslide is numerically studied in the following contents.
The deformation and failure of the Yanguan landslide are primarily governed by intrinsic factors, including geological and topographical attributes of the region. The sliding mass consists of aged, loosely compacted deposits, characterized by relatively low strength, which facilitates water infiltration and migration. Moreover, a seasonal gully near the left boundary of the landslide tends to accumulate surface runoff. Externally, the rainfall and the water level act as triggers for the landslide initiation and ensuing damage. From 18 September to 26 October 2017, the research area experienced a cumulative rainfall of 356.4 mm, significantly higher at 2.5 times the average monthly precipitation over the past decade. Additionally, the TGR typically reaches its peak water level of 175 m in October. As a result, before the Yanguan landslide event, this area was subjected to the combined effects of prolonged, continuous rainfall and a high reservoir water level. This condition saturated a significant portion of the rock and soil mass of the landslide, reducing its stability and prompting movement and damage.

3. Methodology

In this section, we detail the methodology for constructing a three-dimensional numerical model of the landslide in question and describe the calibration process for the modeling parameters. It should be emphasized that the material degradation discussed earlier serves primarily to elucidate the failure mechanism of the landslide. In contrast, the PFC model is specifically designed to simulate the landslide runout behavior. Consequently, material degradation is not factored into the numerical model that has been constructed.

3.1. Numerical Model of the Yanguan Landslide

To simulate the runout dynamics of the Yanguan landslide, the Particle Flow Code (PFC) software [43] has been utilized, and it was selected for its unique capacity to simulate particle-scale fragmentation dynamics and interlocking effects through clump logic—critical for modeling the multi-stage disintegration and energy transfer processes inherent to reservoir landslides under cyclic hydraulic loading—while maintaining computational feasibility via parallelized granular mechanics solvers. In PFC simulations, rigid particles are connected through contact points to form rock masses, and the fragmentation and breakdown of these rock masses are modeled by the failure of these contacts. The interactions among elements in the PFC model represent a dynamic process that seeks equilibrium within the system. During the computational process, Newton’s second law and force-displacement relationships are alternately applied [44], and the contact forces and torques between updated elements are calculated based on force-displacement theory [45]. Using this numerical tool, an initial numerical model of the landslide has been constructed, as illustrated in Figure 6.
Accurate terrain data are crucial for the development of a three-dimensional PFC model of the landslide [46]. In this research, digital elevation model (DEM) data from both pre-slide and post-slide states were utilized to formulate the numerical representation of the landslide. To capture the post-failure topography, a UAV survey was deployed, yielding 676 high-resolution aerial images at a resolution of 0.0282 m per pixel, captured across three flights. Note that the altitude of the UAV is 800 m, and the forward and side overlap percentages are 60% and 30%, respectively. For the processing of these UAV images, the UAV Manager Software, provided by FEIMA Robotics Technology Company (Shenzhen, China), was employed, and the procedures can be organized as follows: first, drone aerial photography data were converted to 3-D point cloud data; then, the point cloud data were categorized into ground points and non-ground points manually; a built-in interpolation method was then adopted for building the digital elevation model (without vegetation) from the ground points. Concurrently, the pre-slide terrain was recreated using a DEM with 10.0 m/pixel resolution, sourced from high-resolution optical remote sensing imagery from Google Earth, predating the landslide event. Following the acquisition of DEMs for both periods, elevation changes were computed, and the landslide volume was quantified through deformation extraction and raster-based volumetric integration. Subsequently, digital elevation data for the slide bed and mass were extracted, the pre-slide and post-slide DEMs were converted to .stl format files, the geometry import command was then used to import geometric data from the files, and a three-dimensional geometric model was assembled, as illustrated in Figure 7. Note that although terrain complexity and vegetation-induced shadowing may introduce localized point cloud uncertainties affecting initial model resolution, the simulated kinematic consistency remains validated through field-verified deposition patterns and debris motion coherence.
In the PFC model of the landslide, the sliding bed and sliding mass are represented by wall and particle elements, respectively, as depicted in Figure 7a. The sliding bed is configured as rigid boundaries, comprising 60,548 facets. The particle sizes are usually based on the tradeoff between computational efficiency and simulation accuracy. For example, the number of particles can increase with the decrease in particle sizes; and the increase in the number of particles could lead to an increase in simulation accuracy but a decrease in computational efficiency. Based on this relationship, the sliding mass is simulated with 13,778 particles, with a radius uniformly ranging from 1.4 to 3.5 m. The interactions between rigid particles, including forces and moments, are simulated using the linear parallel bond contact model. Note that the hydro-mechanical coupling is not considered in this study, which may lead to misjudgment of sliding surface development patterns. Future improvements will incorporate a fully coupled hydro-mechanical framework to better reproduce the runout process of the landslides.

3.2. Determination of the Modeling Parameters of the Built PFC Model

In PFC simulations of landslides, the runout behavior of the sliding mass is significantly influenced by various modeling parameters [47]. These include damping coefficients, contact micro-parameters between particles in the sliding mass and rigid boundaries like the sliding bed, and contact micro-parameters, etc. Directly determining these parameters is quite a difficult issue. Typically, they are established through systematic adjustments or trial-and-error analyses [48], drawing on limited knowledge from previous research, and the calibration results are shown in Table 1.
To delineate the attributes of the sliding mass collected from the Yanguan landslide, the uniaxial compression tests were executed using samples sourced directly from the landslide area. Macro properties such as uniaxial compressive strength (UCS), Young’s modulus (E), and Poisson’s ratio (ν) were utilized as benchmark macro-parameters for the sliding mass. Traditional parameter calibration in particle-based simulations often relies on iterative heuristic approaches that face inherent limitations in addressing complex parameter interdependencies, and it struggles with high-dimensional parameter spaces and nonlinear micro-macro relationships, often requiring thousands of trials to achieve suboptimal solutions. By contrast, the SVM-based approach coupled with uniaxial compression tests leverages kernel functions to decouple parameter interactions and enables multi-objective optimization of stress–strain matching and failure mode consistency. Thereby, this overcomes critical limitations of conventional methods in PFC parameter calibration. In this research, a hybrid approach integrating the support vector machine (SVM) method with uniaxial compression simulations was employed to fine-tune the inter-particle contact micro-parameters, which include the effective modulus (Ec), the ratio of normal to shear stiffness (K), the bond’s normal strength (σc), the bond’s shear strength (τc), and the friction coefficient (μ).
Initially, systematic iterative testing was conducted to determine the statistical distributions of particle contact parameters, ensuring micromechanical calibration aligns with observed macroscopic behaviors. Following this, 80 distinct sets of micro-parameters were algorithmically generated based on the acquired statistical data. Subsequently, three-dimensional uniaxial compression simulations (referenced in Figure 8) were carried out to deduce the associated macro-parameters. A parametric study was carried out in the preliminary analysis, which indicates that the runout behavior is more influenced by the bond strength and the friction coefficient among particles.
It should be noted that although the SVM model was trained with limited samples, it demonstrates robust predictive capability. Initial validation using standard testing datasets revealed strong agreement between predictions and observations. Expanded verification with additional test cases further confirms model reliability. The adopted sample size also aligns with established discrete element simulation practices in parameter calibration studies [47,49]. These results indicate that the selected training sample size was indeed sufficient for the SVM model training in this context.
The calibrated SVM algorithm enables systematic optimization of particle-scale parameters (see Table 1) against benchmark geomechanical properties. Implemented in DEM simulations of uniaxial compression (see Figure 8), these parameters yield mechanical responses demonstrating close alignment with target criteria (R2 > 0.92), thereby validating their applicability in reconstructing the Yanguan landslide numerical framework.

4. Results

4.1. Effectiveness of the Built Numerical Model of the Yanguan Landslide

The calibrated micromechanical parameters, when implemented in the discrete element framework, enable high-fidelity reconstruction of the Yanguan landslide runout behavior, with computational deposition patterns aligning closely with field-mapped extents (see Figure 9). Comparing the simulated deposit with the actual deposit obtained from field surveys and UAV photogrammetry, the spatial extent matching index and volume consistency reveal that the PFC simulation accurately captures the primary runout characteristics of the landslide. Specifically, both the three-dimensional distribution and two-dimensional attributes of the simulated deposit closely match the observed behaviors. The simulated displacement is approximately 76.26 m, which is consistent with the field survey results. Additionally, the peak velocity of 2.48 m/s aligns well with the theoretical calculations [50], thereby validating the effectiveness of the constructed numerical model.
It should be noted that the current PFC simulation does not account for the interaction between the sliding mass and water. Future analyses of reservoir landslide runout behaviors may benefit from a more comprehensive simulation approach that integrates solid-fluid composite materials and landslide-water body interactions.
Through the fish function inside the PFC software, the velocity distribution of the Yanguan landslide at different time intervals can be obtained, as depicted in Figure 9. The landslide’s runout duration is approximately 80 s, consistent with the actual event timeline. At the onset, the leading front initiates motion through gravitational stresses within the sliding mass, transforming elevation-derived gravitational energy into progressive kinematic energy. This results in a continuous increase in velocity, with the front edge moving slightly faster than the middle and rear sections, as shown in Figure 9a,b. The average velocity reaches a peak of approximately 2.48 m/s around 5 s (see Figure 9c). During the deceleration phase (5–10 s), increased particle collisions and friction due to the disintegration of the sliding material cause the landslide velocity to gradually decrease. Simultaneously, the leading edge begins to enter the Xiangxi River (refer to Figure 9d–h). Eventually, the sliding mass accumulates at the base of the landslide and deposits into the Xiangxi River. By around 80 s, most particles stabilize, indicating that the landslide has reached a steady state (as illustrated in Figure 9i). This phenomenon closely aligns with field observations of the landslide deposition phase, which accurately replicates the temporal progression of slope failure, deposition geometry, and kinematic patterns evident in post-event surveys.

4.2. Numerical Analysis of the Yanguan Landslide

To examine the velocity distribution throughout the landslide failure process, nine particles within the sliding mass were monitored (Figure 7a). The mean velocity of the S255 Road at the landslide’s front was also measured. The monitored velocities are shown in Figure 10. Specifically, Figure 10a compares the average velocities of the landslide and the adjacent road. In the PFC simulation, the road section consists of 112 particles, with the road’s average velocity calculated from these particles. The data indicate that the sliding mass’s velocity initially rises gradually, peaks at 2.48 m/s at 5 s, and then declines. The road velocity follows a similar trend, reaching a peak of 2.98 m/s, which exceeds the landslide’s average velocity. Such a velocity discrepancy between the sliding mass (peak 2.48 m/s) and adjacent road section (peak 2.98 m/s) stems from differential energy transmission mechanisms. As the leading-edge particles transferred kinetic energy to the rigid road structure through impact forces, the road’s lower friction coefficient and reduced interparticle collisions enabled sustained acceleration until structural failure at t = 5 s.
As illustrated in Figure 10b, the average velocities at the landslide tail end, midsection, and forward edge were also monitored. The peak velocities varied, ranging from 1.75 m/s to 2.26 m/s at the rear, 4.22 m/s to 4.47 m/s in the middle, and 4.50 to 5.88 m/s at the front. These peaks were recorded between 4 s and 40 s, with the forward edge consistently exhibiting higher peak velocities compared to the rear. Furthermore, the forward edge reached its peak velocity and stabilized more quickly than the rear section. The kinematic divergence across sliding mass segments originates from spatial variations in energy conversion efficiency. Frontal regions achieve peak velocities through unconstrained gravitational drive with transient basal lubrication, while intermediate zones experience velocity damping via interparticle momentum transfer and developing shear plane resistance. Posterior sections display progressive velocity attenuation governed by strain energy redistribution through fragment reorientation and intra-mass collision cascades, with these dynamic gradients becoming exacerbated through motion persistence.
Figure 11 displays the displacement profiles of nine key particles, the road’s average displacement, and the total displacement of the sliding mass. In Figure 11a, the sliding mass and road show average displacements of 76.26 m and 74.59 m, respectively. The rear-edge particles (No. 1–3) in Figure 11b have much shorter displacements and sliding durations. The green lines in Figure 11b track particles No. 4–6, aligning with the landslide’s average displacement. The purple lines show front-edge particles’ displacements, highlighting faster initiation and longer movement durations. Particle No. 6 travels the farthest at 120 m, outdistancing the other five points in the front and middle sections. Such a phenomenon of frontal mobility dominance arises from both terrain slope differences (34° front vs. 28° upslope) amplifying gravitational drive, and progressive material fragmentation generating finer debris (D50 = 0.8 m) that reduces basal friction. PFC simulations confirm slope geometry initiates frontal acceleration while material gradation governs subsequent energy dissipation through shear band development.
To gain deeper insights into the movement process of the Yanguan landslide, a numerical analysis was performed to trace the conversion of different energy forms during the landslide’s runout, while also monitoring crack formations to delineate distinct movement stages. As shown in Figure 12, gravitational potential energy is primarily converted into frictional, collision, and kinetic energies, with strain energy playing a negligible role. Consequently, strain energy is excluded from the analysis.
In the initial 8 s interval, gravitational potential energy is largely converted into frictional and kinetic energies, with collision energy being less significant, and the conversion rate of the gravitational potential energy is 11.68%. The disintegration of the sliding mass, evidenced by a rise in crack numbers (Figure 12), leads to an increased conversion of gravitational potential energy into frictional and collision energies. Crack numbers provide a real-time measure of bond breakages in the landslide’s initial condition. During the first 8 s, the landslide moves and accelerates as a unified body, with minimal internal fracturing, resulting in the primary conversion of gravitational potential energy into frictional and kinetic energies.
During the early acceleration phase (8 s to 20 s), the sliding mass continues to accelerate, leading to a marked increase in kinetic energy, which peaks at 26.37 GJ at 8 s, corresponding to a 4.22% conversion rate. The frictional interactions are more pronounced than collision effects, leading to greater frictional energy dissipation than collision energy dissipation at this stage.
During the deceleration phase (20 s to 42 s), the sliding mass disintegrates faster, bonded particles separate, and particle collisions become more intense, leading to higher collision energy dissipation. At the same time, significant friction and impact at the sliding bed and base cause a notable increase in frictional energy. It should be noted that the late-stage frictional energy increase arises from terrain-driven kinematic transitions: as the landslide front encounters flatter terrain (slope < 15°), debris deceleration converts kinetic energy into frictional dissipation through two mechanisms: (1) enhanced interlocking between basal fragments and bedrock protrusions, evidenced by concentrated contact forces in DEM simulations, and (2) compression-induced particle collisions due to frontal debris accumulation.
After 42 s, the landslide’s kinetic energy gradually decreases, while frictional and collision energy dissipations keep rising until the landslide stabilizes. Overall, frictional and collisional mechanisms dominate energy dissipation, accounting for 296.40 GJ (48.59%) and 172.53 GJ (28.28%) of the total released gravitational potential energy, respectively. Figure 12 also shows that as the velocity and displacement of the sliding mass increase, gravitational potential energy decreases, while frictional and collision energies rise, highlighting the close link between the runout behavior and its energy dissipation process.
Comparative studies in reservoir landslide modeling reveal distinct methodological emphases: While finite element methods (FEM) have been widely applied to analyze seepage-stability coupling under water-level fluctuations [1], their continuum-based formulations struggle to capture post-failure fragmentation processes. Material point method (MPM) implementations, as seen in Oso landslide simulations [30], effectively model large deformations but often simplify interparticle interactions. Smooth particle hydrodynamics (SPH) approaches excel in fluid–solid coupling for impulse wave prediction yet frequently overlook soil arching effects critical in reservoir bank failures [51]. The present DEM framework bridges these gaps by explicitly simulating particle-scale mechanics while maintaining computational feasibility through machine learning-optimized parameterization, particularly advancing understanding of energy redistribution patterns during reservoir drawdown-triggered failures. This contrasts with conventional limit equilibrium analyses that neglect dynamic process evolution altogether.

5. Conclusions

This research undertakes a numerical investigation into the dynamic behavior of the Yanguan landslide, which took place in October 2017 in Zigui County, within TGR region in China. The principal conclusions are as follows:
(1)
The Yanguan landslide was likely triggered by a combination of water level fluctuations in the TGR area and intense rainfall. Reservoir drawdown induced basal erosion and pore pressure disequilibrium, reducing shear resistance at the sliding surface, and rainfall accelerated strength degradation through preferential infiltration pathways, triggering progressive failure propagation. These factors may have weakened the physical and mechanical properties of the sliding mass and potentially caused erosion at the landslide base.
(2)
The spatial velocity gradient observed in the landslide’s sliding mass fundamentally stems from energy redistribution mechanisms: unobstructed gravitational acceleration at the front edge created momentum dominance, while progressive energy dissipation through particle collisions and basal friction moderated middle/rear velocities. This self-reinforcing velocity disparity governs both runout dynamics and final deposition patterns, demonstrating that slope failure evolution in reservoir environments is intrinsically controlled by differential energy partitioning along sliding trajectories. The simulated velocity structure’s alignment with field-observed kinematic signatures (orientation/deposition geometry) validates the model’s capacity to capture essential motion characteristics despite current hydrodynamic simplifications.
(3)
This study enhances landslide risk assessment practices by quantifying dynamic parameters critical for impact prediction: First, energy dissipation patterns showing 55–65% of kinetic energy conversion into basal friction during stabilization phases; second, particle interaction mechanisms explaining preferential flow paths along weak geological interfaces. These findings enable more accurate mobility predictions through DEM-based scenario libraries specifically developed for reservoir bank slopes, while the calibration methodology provides a replicable framework for determining region-specific micro-mechanical parameters using limited monitoring data. Further, the results of this work can provide a scientific basis for the study of this type of landslide in the TGR area and contribute to disaster prevention and mitigation.

Author Contributions

Conceptualization, H.F. and B.L.; methodology, B.L. and H.F.; validation, K.L. and Y.M.; writing—original draft preparation, H.F.; writing—review and editing, B.L. and K.L.; funding acquisition, B.L. and Y.M. All authors have read and agreed to the published version of the manuscript.

Funding

The work was funded by Zhejiang Province Investigation Project on Developmental Patterns of Group-Triggered Slope Debris Flows (No. 2024-02-004), China Institute Geo-Environment Monitoring Research and Development Fund Project (No. 20240102), and the National Key Research and Development Program of China (No. 2023YFC3007202).

Data Availability Statement

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Abbreviations

TGRThree Gorges Reservoir
DEMDigital Elevation Model
PFCParticle Flow Code
UAVUnmanned Aerial Vehicle
aslAbove sea level
UCSUniaxial compressive strength
EYoung’s modulus
νPoisson’s ratio
ρDensity
SVMSupport Vector Machine
RminMinimum particle radius
RmaxMaximum particle radius
ρbParticle density
EcEffective modulus
KNormal/shear stiffness ratio
σcBond’s normal strength
τcBond’s shear strength
Ec′Parallel bond normal strength
K′Parallel bond normal/shear stiffness ratio
μFriction coefficient
R2Correlation coefficient

References

  1. Tang, H.; Wasowski, J.; Juang, C.H. Geohazards in the Three Gorges Reservoir Area, China–Lessons learned from decades of research. Eng. Geol. 2019, 261, 105267. [Google Scholar] [CrossRef]
  2. Gong, W.; Zhang, S.; Juang, C.H.; Tang, H.; Pudasaini, S.P. Displacement prediction of landslides at slope-scale: Review of physics-based and data-driven approaches. Earth-Sci. Rev. 2024, 258, 104948. [Google Scholar] [CrossRef]
  3. Petley, D. Global patterns of loss of life from landslides. Geology 2012, 40, 927–930. [Google Scholar] [CrossRef]
  4. Froude, M.J.; Petley, D.N. Global fatal landslide occurrence from 2004 to 2016. Nat. Hazards Earth Syst. Sci. 2018, 18, 2161–2181. [Google Scholar] [CrossRef]
  5. Gong, W.; Juang, C.H.; Wasowski, J. Geohazards and human settlements: Lessons learned from multiple relocation events in Badong, China–Engineering geologist’s perspective. Eng. Geol. 2021, 285, 106051. [Google Scholar] [CrossRef]
  6. Yin, Y.; Huang, B.; Wang, W.; Wei, Y.; Ma, X.; Ma, F.; Zhao, C. Reservoir-induced landslides and risk control in Three Gorges Project on Yangtze River, China. J. Rock Mech. Geotech. Eng. 2016, 8, 577–595. [Google Scholar] [CrossRef]
  7. Wang, J.; Xiao, L.; Zhang, J.; Zhu, Y. Deformation characteristics and failure mechanisms of a rainfall-induced complex landslide in Wanzhou County, Three Gorges Reservoir, China. Landslides 2020, 17, 419–431. [Google Scholar] [CrossRef]
  8. Zhang, Y.; Hu, X.; Tannant, D.D.; Zhang, G.; Tan, F. Field monitoring and deformation characteristics of a landslide with piles in the Three Gorges Reservoir area. Landslides 2018, 15, 581–592. [Google Scholar] [CrossRef]
  9. Zhang, J.; Tang, H.; Tan, Q.; Mao, M.; Zhou, B.; Zhang, Y. A generalized early warning criterion for the landslide risk assessment: Deformation probability index (DPI). Acta Geotech. 2024, 19, 2607–2627. [Google Scholar] [CrossRef]
  10. Zhang, J.; Tang, H.; Zhou, B.; Wen, T.; Zhang, S. A new early warning criterion for landslides movement assessment: Deformation Standardized Anomaly Index. Bull. Eng. Geol. Environ. 2024, 83, 205. [Google Scholar] [CrossRef]
  11. Huang, D.; Gu, D.; Song, Y.; Cen, D.; Zeng, B. Towards a complete understanding of the triggering mechanism of a large reactivated landslide in the Three Gorges Reservoir. Eng. Geol. 2018, 238, 36–51. [Google Scholar] [CrossRef]
  12. Li, Y.; Xie, X.; Jin, B.; Chen, L.; Liang, X.; Yin, K. Comprehensive risk management of reservoir landslide-tsunami hazard chains: A case study of the Liangshuijing landslide in the Three Gorges Reservoir area. Landslides 2024, 22, 671–691. [Google Scholar] [CrossRef]
  13. Du, W.; Sheng, Q.; Fu, X.; Chen, J.; Zhou, Y. A TPDP-MPM-based approach to understanding the evolution mechanism of landslide-induced disaster chain. J. Rock Mech. Geotech. Eng. 2022, 14, 1200–1209. [Google Scholar] [CrossRef]
  14. Gu, D.; Huang, D.; Yang, W.; Zhu, J.; Fu, G. Understanding the triggering mechanism and possible kinematic evolution of a reactivated landslide in the Three Gorges Reservoir. Landslides 2017, 14, 2073–2087. [Google Scholar] [CrossRef]
  15. Song, K.; Wang, F.; Yi, Q.; Lu, S. Landslide deformation behavior influenced by water level fluctuations of the Three Gorges Reservoir (China). Eng. Geol. 2018, 247, 58–68. [Google Scholar] [CrossRef]
  16. Xu, S.; Niu, R. Displacement prediction of Baijiabao landslide based on empirical mode decomposition and long short-term memory neural network in Three Gorges area, China. Comput. Geosci. 2018, 111, 87–96. [Google Scholar] [CrossRef]
  17. Ma, J.; Tang, H.; Liu, X.; Wen, T.; Zhang, J.; Tan, Q.; Fan, Z. Probabilistic forecasting of landslide displacement accounting for epistemic uncertainty: A case study in the Three Gorges Reservoir area, China. Landslides 2018, 15, 1145–1153. [Google Scholar] [CrossRef]
  18. Kang, X.; Xu, G.; Yu, Z.; Wang, S.; Wang, M. Experimental investigation of the interaction between water and shear-zone materials of a bedding landslide in the Three Gorges Reservoir Area, China. Bull. Eng. Geol. Environ. 2020, 79, 4079–4092. [Google Scholar] [CrossRef]
  19. Miao, F.; Wu, Y.; Török, Á.; Li, L.; Xue, Y. Centrifugal model test on a riverine landslide in the Three Gorges Reservoir induced by rainfall and water level fluctuation. Geosci. Front. 2022, 13, 101378. [Google Scholar] [CrossRef]
  20. Yin, Y.; Huang, B.; Chen, X.; Liu, G.; Wang, S. Numerical analysis on wave generated by the Qianjiangping landslide in Three Gorges Reservoir, China. Landslides 2015, 12, 355–364. [Google Scholar] [CrossRef]
  21. Li, B.; Tang, H.; Gong, W.; Cheng, Z.; Li, T.; Wang, L. Numerical study of the runout behavior of the Kamenziwan landslide in the Three Gorges Reservoir region, China. Landslides 2022, 19, 963–976. [Google Scholar] [CrossRef]
  22. Li, B.; Gong, W.; Tang, H.; Zou, Z.; Bowa, V.M.; Juang, C.H. Probabilistic analysis of a discrete element modelling of the runout behavior of the Jiweishan landslide. Int. J. Numer. Anal. Methods Geomech. 2021, 45, 1120–1138. [Google Scholar] [CrossRef]
  23. Gong, W.; Zhao, C.; Juang, C.H.; Zhang, Y.; Tang, H.; Lu, Y. Coupled characterization of stratigraphic and geo-properties uncertainties–a conditional random field approach. Eng. Geol. 2021, 294, 106348. [Google Scholar] [CrossRef]
  24. Zhao, C.; Gong, W.; Li, T.; Juang, C.H.; Tang, H.; Wang, H. Probabilistic characterization of subsurface stratigraphic configuration with modified random field approach. Eng. Geol. 2021, 288, 106138. [Google Scholar] [CrossRef]
  25. Zhang, J.; Lin, C.; Tang, H.; Wen, T.; Tannant, D.D.; Zhang, B. Input-parameter optimization using a SVR based ensemble model to predict landslide displacements in a reservoir area—A comparative study. Appl. Soft Comput. 2024, 150, 111107. [Google Scholar] [CrossRef]
  26. Luo, S.L.; Jin, X.G.; Huang, D. Long-term coupled effects of hydrological factors on kinematic responses of a reactivated landslide in the Three Gorges Reservoir. Eng. Geol. 2019, 261, 105271. [Google Scholar] [CrossRef]
  27. Zhang, J.; Chen, C.; Wu, C.; Zhang, Y. Development of An Image-based Borehole Flowmeter for Real-time Monitoring of Groundwater Flow Velocity and Direction in Landslide Boreholes. IEEE Sens. J. 2024, 24, 42079–42087. [Google Scholar] [CrossRef]
  28. Pudasaini, S.P.; Mergili, M. A multi-phase mass flow model. J. Geophys. Res. Earth Surf. 2019, 124, 2920–2942. [Google Scholar] [CrossRef]
  29. Wang, Q.; Xing, A.; Liao, L.; Liu, Y.; Zhuang, Y. Insights into small landslides inducing major disasters in the Wumeng Mountain area from the Liangshui landslide. Landslides 2024, 22, 857–875. [Google Scholar] [CrossRef]
  30. Yerro, A.; Soga, K.; Bray, J. Runout evaluation of Oso landslide with the material point method. Can. Geotech. J. 2019, 56, 1304–1317. [Google Scholar] [CrossRef]
  31. Zhang, T.; Yin, Y.; Li, B.; Liu, X.; Wang, M.; Gao, Y.; Wan, J.; Gnyawali, K.R. Characteristics and dynamic analysis of the February 2021 long-runout disaster chain triggered by massive rock and ice avalanche at Chamoli, Indian Himalaya. J. Rock Mech. Geotech. Eng. 2023, 15, 296–308. [Google Scholar] [CrossRef]
  32. Hungr, O.; Leroueil, S.; Picarelli, L. The Varnes classification of landslide types, an update. Landslides 2014, 11, 167–194. [Google Scholar] [CrossRef]
  33. Hungr, O.; McDougall, S. Two numerical models for landslide dynamic analysis. Comput. Geosci. 2009, 35, 978–992. [Google Scholar] [CrossRef]
  34. Li, B.; Tang, H.; Gong, W. Probabilistic analysis of post-failure runout behavior of landslides. In Proceedings of the 7th International Symposium on Geotechnical Safety and Risk (ISGSR 2019), Taipei, Taiwan, 11–13 December 2019; pp. 11–13. [Google Scholar]
  35. Li, B.; Gong, W.; Tang, H.M.; Wang, L. DEM Simulation of the Bridge Collapse under the Impact of Rock Avalanche: A Case Study of the 2020 Yaoheba Rock Avalanche in Southwest China. Bull. Eng. Geol. Environ. 2024, 83, 104. [Google Scholar] [CrossRef]
  36. Wang, H.; Liu, S.; Xu, W.; Yan, L.; Qu, X.; Xie, W. Numerical investigation on the sliding process and deposit feature of an earthquake-induced landslide: A case study. Landslides 2020, 17, 2671–2682. [Google Scholar] [CrossRef]
  37. Wei, L.; Cheng, H.; Dai, Z. Propagation Modeling of Rainfall-Induced Landslides: A Case Study of the Shaziba Landslide in Enshi, China. Water 2023, 15, 424. [Google Scholar] [CrossRef]
  38. Sun, L.; Shen, F.; Li, C. Water-induced weakening retrogressive landslides: A study on the progressive failure process and formation mechanism of trailing edge fracture surfaces. Bull. Eng. Geol. Environ. 2023, 82, 201. [Google Scholar] [CrossRef]
  39. Zhu, D.; Song, K.; Mu, J.; Huang, H.; Du, H.; Xiazhuang, F.; Sun, C. Effect of climate change induced extreme precipitation on landslide activity in the Three Gorges Reservoir, China. Bull. Eng. Geol. Environ. 2021, 80, 781–794. [Google Scholar] [CrossRef]
  40. Wu, Q.; Liu, Y.; Tang, H.; Kang, J.; Wang, L.; Li, C.; Wang, D.; Liu, Z. Experimental study of the influence of wetting and drying cycles on the strength of intact rock samples from a red stratum in the Three Gorges Reservoir area. Eng. Geol. 2023, 314, 107013. [Google Scholar] [CrossRef]
  41. Yin, Y.; Huang, B.; Zhang, Q.; Yan, G.; Dai, Z. Research on recently occurred reservoir-induced Kamenziwan rockslide in Three Gorges Reservoir, China. Landslides 2020, 17, 1935–1949. [Google Scholar] [CrossRef]
  42. Yue, X.; Wang, Y.; Wen, T. An Inversion Study of Reservoir Colluvial Landslide Permeability Coefficient by Combining Physical Model and Data-Driven Models. Water 2024, 16, 686. [Google Scholar] [CrossRef]
  43. ICG (Itasca Consulting Group). PFC—Particle Flow Code Documentation, version 6.0; Itasca Consulting Group Inc.: Minneapolis, MN, USA, 2020.
  44. Castro-Filgueira, U.; Alejano, L.R.; Arzúa, J.; Ivars, D.M. Sensitivity analysis of the micro-parameters used in a PFC analysis towards the mechanical properties of rocks. In Proceedings of the ISRM EUROCK, Ostrava, Czech Republic, 20–22 June 2017. [Google Scholar]
  45. Potyondy, D.O.; Cundall, P.A. A bonded-particle model for rock. Int. J. Rock Mech. Min. Sci. 2007, 41, 1329–1364. [Google Scholar] [CrossRef]
  46. Cheng, Z.; Gong, W.; Tang, H.; Juang, C.H.; Deng, Q.; Chen, J.; Ye, X. UAV photogrammetry-based remote sensing and preliminary assessment of the behavior of a landslide in Guizhou, China. Eng. Geol. 2021, 289, 106172. [Google Scholar] [CrossRef]
  47. Wei, J.; Zhao, Z.; Xu, C.; Wen, Q. Numerical investigation of landslide kinetics for the recent Mabian landslide (Sichuan; China). Landslides 2019, 16, 2287–2298. [Google Scholar] [CrossRef]
  48. Calvello, M.; Cuomo, S.; Ghasemi, P. The role of observations in the inverse analysis of landslide propagation. Comput. Geotech. 2017, 92, 11–21. [Google Scholar] [CrossRef]
  49. Fathipour-Azar, H. Machine learning-assisted distinct element model calibration: ANFIS, SVM, GPR, and MARS approaches. Acta Geotech. 2021, 17, 1207–1217. [Google Scholar] [CrossRef]
  50. Hungr, O.; Evans, S.G. Entrainment of debris in rock avalanches: An analysis of a long run-out mechanism. Geol. Soc. Am. Bull. 2004, 116, 1240–1252. [Google Scholar] [CrossRef]
  51. Dai, Z.; Huang, Y.; Xu, Q. A hydraulic soil erosion model based on a weakly compressible smoothed particle hydrodynamics method. Bull. Eng. Geol. Environ. 2019, 78, 5853–5864. [Google Scholar] [CrossRef]
Water 17 00695 g001
Figure 1. Geological background of the research area: (a) Geomorgraphy of the Zigui County; (b) Lithologic map of the study area.
Figure 1. Geological background of the research area: (a) Geomorgraphy of the Zigui County; (b) Lithologic map of the study area.
Water 17 00695 g001
Water 17 00695 g002
Figure 2. Images taken after the Yanguan landslide occurrence: (a) Post-failure form of the landslide; (b) Destroyed house; (c) Ruined road; (d) Cracks.
Figure 2. Images taken after the Yanguan landslide occurrence: (a) Post-failure form of the landslide; (b) Destroyed house; (c) Ruined road; (d) Cracks.
Water 17 00695 g002
Water 17 00695 g003
Figure 3. Geological profile of the Yanguan landslide.
Figure 3. Geological profile of the Yanguan landslide.
Water 17 00695 g003
Water 17 00695 g004
Figure 4. Water level variation and rainfall distribution in YGR from 2003 to 2019.
Figure 4. Water level variation and rainfall distribution in YGR from 2003 to 2019.
Water 17 00695 g004
Water 17 00695 g005
Figure 5. Morphology and movement deposits characteristics of the Yanguan landslide.
Figure 5. Morphology and movement deposits characteristics of the Yanguan landslide.
Water 17 00695 g005
Water 17 00695 g006
Figure 6. Numerical modeling flowchart of the PFC model of the Yanguan landslide.
Figure 6. Numerical modeling flowchart of the PFC model of the Yanguan landslide.
Water 17 00695 g006
Water 17 00695 g007
Figure 7. Numerical model of the landslide: (a) PFC model; (b) Terrain and stratum distribution.
Figure 7. Numerical model of the landslide: (a) PFC model; (b) Terrain and stratum distribution.
Water 17 00695 g007
Water 17 00695 g008
Figure 8. A 3-D uniaxial compression test of the sliding mass using PFC.
Figure 8. A 3-D uniaxial compression test of the sliding mass using PFC.
Water 17 00695 g008
Water 17 00695 g009
Figure 9. Distribution of the Yanguan landslide velocity: (a) t = 0 s; (b) t = 10 s; (c) t = 20 s; (d) t = 30 s; (e) t = 40 s; (f) t = 50 s; (g) t = 60 s; (h) t = 70 s; (i) t = 80 s.
Figure 9. Distribution of the Yanguan landslide velocity: (a) t = 0 s; (b) t = 10 s; (c) t = 20 s; (d) t = 30 s; (e) t = 40 s; (f) t = 50 s; (g) t = 60 s; (h) t = 70 s; (i) t = 80 s.
Water 17 00695 g009
Water 17 00695 g010
Figure 10. Average velocities of the Yanguan landslide: (a) average velocities of the road and sliding mass; (b) average velocities monitored at different locations of landslide.
Figure 10. Average velocities of the Yanguan landslide: (a) average velocities of the road and sliding mass; (b) average velocities monitored at different locations of landslide.
Water 17 00695 g010
Water 17 00695 g011
Figure 11. Average displacements of the Yanguan landslide: (a) average displacements of the road and sliding mass; (b) average displacements monitored at different locations of landslide.
Figure 11. Average displacements of the Yanguan landslide: (a) average displacements of the road and sliding mass; (b) average displacements monitored at different locations of landslide.
Water 17 00695 g011
Water 17 00695 g012
Figure 12. Energy conversions during the landslide runout process.
Figure 12. Energy conversions during the landslide runout process.
Water 17 00695 g012
Table 1. Modeling parameters adopted in the Yanguan landslide model.
Table 1. Modeling parameters adopted in the Yanguan landslide model.
Particle–Particle Contact Micro-ParametersMacro-Parameters
ItemValueItemValue
Minimum particle radius, Rmin (m)1.40Density, ρ (kg/m3)2750.00
Maximum particle radius, Rmax (m)3.50
Particle density, ρb (kg/m3)2750.00Uniaxial compressive strength, UCS (MPa)3.26
Effective modulus, Ec (GPa)1.80
Normal/shear stiffness ratio, K2.00Young’s modulus, E (GPa)2.45
Parallel bond normal strength, σc (MPa)3.60
Parallel bond shear strength, τc (MPa)3.60
Parallel bond modulus, Ec′ (GPa)1.80Poisson’s ratio, ν0.26
Parallel bond normal/shear stiffness ratio, K′2.00
Friction coefficient, μ0.40
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Fang, H.; Li, B.; Liu, K.; Meng, Y. Numerical Investigation into the Runout Dynamics of Reservoir Landslides: Insights from the Yanguan Landslide. Water 2025, 17, 695. https://doi.org/10.3390/w17050695

AMA Style

Fang H, Li B, Liu K, Meng Y. Numerical Investigation into the Runout Dynamics of Reservoir Landslides: Insights from the Yanguan Landslide. Water. 2025; 17(5):695. https://doi.org/10.3390/w17050695

Chicago/Turabian Style

Fang, Hao, Bing Li, Kai Liu, and Yaobin Meng. 2025. "Numerical Investigation into the Runout Dynamics of Reservoir Landslides: Insights from the Yanguan Landslide" Water 17, no. 5: 695. https://doi.org/10.3390/w17050695

APA Style

Fang, H., Li, B., Liu, K., & Meng, Y. (2025). Numerical Investigation into the Runout Dynamics of Reservoir Landslides: Insights from the Yanguan Landslide. Water, 17(5), 695. https://doi.org/10.3390/w17050695

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop