Deformation and Failure Mechanism of a Massive Ancient Anti-Dip River-Damming Landslide in the Upper Jinsha River

Abstract

Landslides are a typical geological hazard that can cause large numbers of casualties and huge economic losses, and the overflow of a weir from a blocked river landslide can have even more disastrous consequences. Of the different types of landslides, about 33% of landslides happen in anti-dip slopes. This paper reports a massive ancient anti-dip river-damming landslide on the Jinsha River: the Zongrongcun landslide. Field investigation and theoretical analysis were used to reveal the potential mechanism of this ancient landslide, and the block discrete element software 3DEC was used to replicate its landslide process. The findings from the present study are as follows: (1) blocks in this landslide were classified into significant slide, significant toppling, and significant slide categories based on D_f. (2) The whole landslide was divided into significant sliding and toppling zones by D_f = 0.5. (3) The results show that the river-damming landslide was likely to be triggered by river erosion, heavy rainfall, gravity. Under strong valley trenching, the rocks on the slope fractured under gravity and tectonic stress. These factors caused rock blocks tensile fracture failure. Then a penetrating sliding surface formed on the slope, which subsequently caused this river-damming landslide.

1. Introduction

As a type of geological hazard, landslides can cause large numbers of casualties and huge economic losses [1,2,3,4]. Such occurrences are of great possibility for large-scale landslides in alpine valley areas that rush into the river valley and cause river damming [5]. This can cause disastrous consequences after the dam break [6,7,8]. For instance, the catastrophic Baige landslide occurred on 10 October 2018 in the Tibet Autonomous Region, China, and formed a barrier lake with a direct economic loss of nearly USD 1 billion (Tian et al., 2020) [9]. In 2014, a landslide occurred near Oso, Washington, USA, causing more than 40 casualties [10].

Large-scale river-damming landslides have occurred on several rivers in China [11,12,13]. Categories of these slopes are different, and their failure methods are also different. However, these landslides were mainly triggered by earthquakes and heavy rainfall [14]. Among different types of landslides, about 33% of landslides occur in anti-dip slopes [15]. Generally, it is considered that the anti-dip slope is stable, or it only causes small-scale damage in the shallow part. However, in the past 20 years, researchers have continuously discovered examples of large-scale anti-dip landslides in the alpine regions in western China. The depth of such landslides often reaches 200 m~300 m [16]. Therefore, the landslide of the anti-dipping slope has the characteristics of large volume and it has a large impact area. Once landslides occur, they potentially have a huge impact on major projects and key roads.

At present, many methods, including analytical methods, physical model tests, and numerical simulations have been used to investigate the mechanism of anti-dipping landslides. In terms of analytical methods, it is not practically possible to evaluate anti-dip failure by limit equilibrium approaches, though limit equilibrium method was widely used because, for anti-dip failure, the resistance and shear stress embattled at the base surface do not play any role in mass stability [17]. Therefore, many researchers evaluated this type of failure through a kinematic analysis method [18,19,20,21]. Amini, Majdi, and Aydan [22] analyzed the flexural-toppling failure based on classification from [23]. Then Amini et al. [24], considering two different states, proposed a close approximation computational approach to analyze slippery single-column and potential block-flexural toppling. These methods are useful in understanding the failure mechanism of anti-dip slopes; however, careful caution is necessary in applying these methods since a number of assumptions are made. In addition, these methods cannot be used to describe the progressive evolution and failure behavior of a landslide. Physical model tests provide an efficient way to observe the movement process directly, but the model tests size of a slope is usually limited. In addition, expense and time are also limitations in physical model tests.

Numerical simulation is considered an effective method that can study the movement process of landslides triggered by anti-dip failure. Many researchers used numerical methods to investigate mechanism of anti-dip landslides [25,26,27,28,29,30]. For example, Azarafza, M. [31] assessed stability of a rock slope through block theory and numerical simulation. Then, in 2020, Azarafza, M [32], based on analytical stability method, studied the key-block for discontinuous rock slopes subjected to anti-dip failure. To date, the numerical research on anti-dip failure mainly focuses on the description of engineering geological phenomena [33,34,35,36] and mechanism and deformation characteristics. According to the different failure phenomena and mechanism and deformation characteristics, the failure forms of toppling landslides have been systematically studied [15,23,37,38,39,40]. However, few works have been conducted to model the whole movement process of landslides triggered by anti-dip failures to sudden failure and final deposition.

In addition, the failure mechanism of ancient landslides is complicated, and its failure may not only be triggered by a single factor. An ancient landslide was usually triggered by multiple factors, including river erosion, heavy rainfall, weight stress, horizontal tectonic stress, etc. Therefore, a whole movement process numerical simulation can better reveal the mechanisms and characteristics of anti-dip landslides. (Methods applied on stability for anti-dip slopes can be seen in Table 1).

This paper reports a study of a large-scale anti-dip ancient landslide in the upper Jinsha River: the Zongrongcun landslide. Detailed field investigation, theoretical analysis, and numerical investigation were conducted to reveal the potential mechanism of the landslide. Numerical simulations using the DEM method are conducted to model the whole movement process of the landslide from anti-dip deformation to sudden failure and formation of a landslide dam due to rock avalanche deposition.

2. Geological Setting

The study area (Figure 1) is located at the southeast Tibet, China, approximately 82 km south from Benzilan Town, Yunnan Province. The geomorphology is steep hillslopes with slope angles varying from 30° to 65° and deeply incised valleys.

Rocks exposed in this region mainly consist of holocene landslide accumulation (Qh^del), holocene alluvial and residual slope deposits (Qh^esl+al), holocene lacustrine sediments, Triassic quartz diorite $({\mathrm{T}}_3\mathsf{\delta }\mathrm{o})$, Mesoproterozoic mica gneiss (Pt₂X³), and Paleozoic Ophiolite (Plagioamphibolite). Affected by tectonic movement in this area, the rock stratum is relatively sheared and tensile damaged. The distribution of stratigraphic lithology is shown in Figure 2. Because the Zongrongcun landslide is a river blocking landslide, there are not only landslide deposits on the river’s left bank but also a large number of Q_h-del Holocene landslide deposits on the river’s right bank. Upstream of the landslide, we found Q4 Holocene lacustrine deposits (mainly composed of silty sand and silt). According to previous studies, the Zengdatong fault, the Lifu-Riyu fault, the Langzhong fault, and the Palaeoscopy fault in the Jinshajiang fault zone showed obvious signs of activity during the late Pleistocene and Holocene. The motion properties are dominated by right-handed strike-slip, with a thrust sliding component. The Holocene right-handed horizontal slip rate is 3.5–4.3 mm/a, and the vertical slip rate is 0.9–1.1 mm/a. The fault has been active from the Epipleistocene Epoch to the Holocene Epoch [41].

There are large number of faults in the study area (Figure 1), which have an important influence on the stability of the slope in this area. The Lifu-Riyu north–south boundary fault (hereinafter referred to as the fault F5b) runs through the Zongrongcun landslide. The fault F5b, forming the boundary between mica schist and ophiolite, has its strike as 7° and its dip angle as about 45°.

The region has a semi-arid climate, with precipitation concentrated from June to September. The average annual precipitation is 468.3 mm in this area. Arid climate and large temperature difference between day and night lead to poor vegetation development [11]. Therefore, the scarcity of vegetation and concentrated heavy rainfall are very detrimental to the stability of the slope. This unfavorable factor is coupled with the erosion of the slope toe from the river, which can easily lead to the occurrence of landslides.

3. Methods

3.1. Landslide Description

The Zongrongcun landslide is located on the left bank of the Jinsha River (Figure 2), with a toe elevation of 2145 m and a headscarp elevation of 2895 m. It covers a surface area of 1.5 km² with a width of 1000 m, a length of 1500 m, and an estimated volume of 4.3 × 10⁷ m³. The toe and headscarp of the landslide are steep (nearly 75°), while the middle part is relatively flat (about 35°). The entire landslide has an obvious chair-like landform. In addition, the left and right sides of the landslide were deeply incised by valleys, forming two huge gullies.

Rocks exposed in landslide area are mainly consist of Paleozoic Ophiolite (Plagioamphibolite), Mesoproterozoic mica schist (Pt₂X³) and Triassic quartz diorite $({\mathrm{T}}_3\mathsf{\delta }\mathrm{o})$. The physical and mechanical properties of the rocks are shown in Table 2.

According to the mineral identification results, the Paleozoic Ophiolite (Plagioamphibolite), which has developing a nematoblastic texture and a schistose structure, mainly consists of hornblende (Hbl) and plagioclase (Pl). The mica schist mainly consists of muscovite, quartz, plagioclase, and biotitle with a lepidoblastic texture and a schistose structure. The quartz diorite $({\mathrm{T}}_3\mathsf{\delta }\mathrm{o})$ mainly consists of plagioclase (Pl), potassium feldspar, and quartz (Qtz) and biotite (Bt), which has a flaky, granular metamorphic structure, a variable granite structure, a primary mylonitized fabric, and a massive structure (Figure 3).

Field investigation shows that mica schists are mostly layered and massive structures. The occurrence of schistosity plane is 73°∠68°. The diorite rock mass is relatively intact, with two sets of joints and an overall block structure. The joint occurrences are 84°∠67° and 30°∠75°. The Paleozoic Ophiolite (Plagioamphibolite) is relatively broken, with well-developed joints, and the rock mass structure is fragmented or scattered. The typical joint occurrences are 13°∠47°, 270°∠38° and 65°∠82°, and there is a structural surface on the headscarp of the landslide, and its occurrence is 264°∠47°. In addition, according to regional geological data, residual horizontal tectonic stress remains in this area.

3.2. Landslide Numerical Replication

To better reveal the potential mechanism of the Zongrongcun landslide, 3DEC5.20, a block discrete element software, was used to replicate this landslide.

The study area is characterized by rock slopes with large number of structural planes, which largely determine the stability of these slopes. Therefore, it is very suitable for 3DEC5.20 to simulate this landslide process [43]. According to field investigation, possible landform before the landslide has been restored to build this numerical model. The model size and structural surface settings are shown in Figure 4. Detailed information for structural planes is shown in Table 3 and Table 4. In 3DEC5.2 software (generated by Itasca, Minneapolis, MN, USA), in order to achieve a more realistic rock fracture, the Burgers-creep viscoplastic model is used for the block, and the constitutive criterion for discontinuities is the Mohr–Coulomb model. The mechanical parameters of rock blocks were obtained through laboratory tests. The strength parameters of joint surface refer to the in situ test results of similar rocks near the dam site of Batang Hydropower Station. The structural plane parameters can be obtained after conversion.

This numerical investigation aims to replicate the landslide process caused by rainfall-coupled river erosion. To consider soften effect of rainfall, we reduce the strength of the structural planes. That is to say, the strength of the structural planes in Table 4 is the strength when the landslide occurred in DEM model. It is worth mentioning that, for an ancient landslide, it is usually impossible to determine precise value of those factors. However, there are still some factors can be determined. The effect of river erosion can be simulated because the Jinsha River has obvious erosion effects on the toe of the slope, and it was also one of the most important factors that conducted many kinds of landslides. In addition, the softening effect of rainfall on the strength of rock structures cannot be ignored in the simulation.

Therefore, in this numerical investigation, two factors were considered, and these factors usually play important roles in other landslides. The major factor considered in this simulation was the effect of river erosion. We simplified the river erosion process into two times. Coupled with the softening effect from rainfall, three cases were simulated in this paper (Table 5). Cases 1 and 3 were carried out in three steps, and Case 2 was divided into two steps.

This simulation considers the deformation and failure process of the Zongrong landslide under three working conditions. The specific working conditions are shown in Table 5. And the river erosion schematic diagram can be found in Figure 5.

(1) Working condition 1 (rainfall infiltration softening): Rainfall mainly affects the strength of the rock structural plane, and in the simulation, the strength of the rock structural plane is reduced by reducing the strength of the rock structural plane.

(2) Working condition 2 (river undercut): In order to simulate the effect of the river erosion on slope toe, the model is realized by excavating the part of the toe of the slope.

(3) Working condition 3 (rainfall + river cut): the combined action of working condition 1 and working condition 2.

4. Results and Discussion

4.1. Numerical Model Verification

A numerical model must be compared with actual field displacement to be validated. However, the landslide in this manuscript is an ancient landslide, which means it is impossible to get the actual field displacement when the landslide occurred. Fortunately, during the process of field investigation, we found some evidence that can speculate where the landslide body slipped to.

Firstly, on the opposite bank and upstream of the landslide, we found we found multiple sites with lacustrine deposits (Figure 2a). These lacustrine sediment samples, after laboratory identification, were formed in a relatively close age. However, no similar lacustrine deposits were found near the downstream of the landslide. Therefore, it can be preliminarily determined that a landslide occurred here. These lacustrine deposits were formed after landslides blocked the river.

Secondly, according to the height of the lacustrine deposition (mentioned in Section 4.4) on the opposite bank of the landslide, the accumulation height of the landslide body can be preliminarily estimated. The accumulation height of the landslide body in the numerical model of this paper is consistent with the lacustrine deposition height. Although there is no actual field displacement, it can still be considered that the numerical model has some certain credibility.

4.2. Results Description

In the landslide area, the 50 monitoring points were divided into five parts from the toe of the landslide to headscarp of the landslide (Figure 6).

Simulation results in Case 1 and Case 2 showed that single factor can be difficult to cause this landslide (Figure 7). The landslide occurred under the effect of rainfall and river erosion. In Case 3, due to the erosion of the river, under the continuous action of rainfall and gravity, the toe of the slope undergoes creep deformation firstly (Figure 7c). The free surface at the toe of the slope increases, the rock mass at the toe of the slope was bent and deformed, and the shallow fracture occurs near the toe of the slope. Subsequently, the toe of the slope was dumped and deformed, and the entire rock formation at the toe of the slope was bent and fractured. Finally, a continuous sliding surface was gradually formed. The trailing edge of the slope was further deformed, eventually forming the large-scale, river-damming Zongrongcun landslide.

Goodman and Bray [23] believed that the toppling deformation of layered slope rock mass could be summarized as flexural toppling, block toppling, and block bending toppling, and they proposed the concept of secondary toppling. Hoek and Bray [44] classified the types of secondary toppling on basis of Goodman and Bray. To understand the failure type of landslide in this DEM model, and to reveal the landslide mechanism, a coefficient has been defined in the following sections.

To describe the toppling and slip degree for each block when landslide occurred, a slip-toppling state coefficient D_f is defined in Equation (1).

(1)$D_f=\frac{V_B}{V_T}$

In Equation (1): D_f is the slip-toppling state coefficient. V_T is the average velocity during the landslide process for a block at point T. V_B is the average velocity during the landslide process for a block at point B, and point T means that it is a point at the top of a block, while point B means that it is a point at the bottom of a block.

When D_f = 1, the block only slides; When 0 < D_f < 1, the block is in the state of slip and toppling. When D_f = 0, only toppling occurs. When 0 < D_f < 1, the smaller the value is, the more the block is inclined to the toppling side.

Fifty monitoring points were set up in the landslide area of the DEM model to monitor the movement characteristics of these blocks (Figure 6). Each monitoring point can obtain a coefficient defined in Equation (1).

If the D_f coefficient for blocks in an area are all more than 0.5, we call this area a significant slip area, and if the D_f coefficient for blocks in an area are all less than 0.5, we call this area a significant toppling area.

It can be seen from Figure 8 that when the landslide occurred, these rock blocks have both slip deformations and toppling deformations. The blocks at toe of the landslide are closest to the slip deformation, and the blocks near the F5b fault are the closest to the toppling deformation, and the block at the headscarp of the landslide is closest to the slip deformation.

According to monitoring data from monitoring points in Figure 6, the displacement curves of the five parts in the landslide process are plotted in Figure 9.

4.3. Potential Mechanism of Toppling

In this work, a simplified cantilever beam model was used to reveal the potential failure mechanism of the landslide. For a block with height Hj and width t of the landslide, it can be simplified as a cantilever beam, and its stress diagram is shown in Figure 8.

In Figure 10, P_j+1 and P_j are the reaction forces of upper and lower adjacent blocks, respectively; T_j+1 and T_j are the friction force generated by sliding between blocks, respectively; h_j+1 and h_j are the distance from P_j+1 and P_j action point to block root, respectively; β is the angle between the block and the horizontal plane; W is block weight.

When the block only produces a toppling deformation, the block rotates around its root, and the bending moment can be expressed in two parts. One is the sliding bending moment, and the other is the resistance bending moment, which are denoted as M_T and M_R, respectively. The expressions are as follows:

(2)$M_T=P_{j+1}·{\mathrm{h}}_{\mathrm{j}+1}+\mathrm{W}·\frac{h}{2}·\cos \mathsf{\beta }$

(3)$M_R=P_j·{\mathrm{h}}_{\mathrm{j}}+\frac{\mathrm{t}}{2}·\left(T_{j+1}+T_j\right)$

According to the Mohr–Coulomb strength criterion, P_j+1, P_j and T_j+1, T_j satisfy:

(4)$T_{j+1}=P_{j+1}·\tan \mathsf{\phi }+{\mathrm{C}}_{\mathrm{j}+1}$

(5)$T_j=P_j·tan\phi +{\mathrm{C}}_{\mathrm{j}}$

$\mathsf{\phi }$ is internal friction angle of rock, C_j+1, C_j is cohesion.

Equations (2) and (3) can obtain the net bending moment of rock block, denoted as M, and then substitute it into Equations (4) and (5) to obtain:

(6)$M=\left({\mathrm{h}}_{\mathrm{j}+1}-\frac{\mathrm{t}}{2}\tan \mathsf{\phi }\right)P_{j+1}-\left(\frac{\mathrm{t}}{2}\tan \mathsf{\phi }+{\mathrm{h}}_{\mathrm{j}}\right)P_j+\frac{\mathrm{h}}{2}\mathrm{Wcos}\mathsf{\beta }-\left({\mathrm{C}}_{\mathrm{j}+1}+{\mathrm{C}}_{\mathrm{j}}\right)$

The tensile stress at the root of the block under the net bending moment is:

(7)${\sigma }_T=\frac{M·\mathrm{y}}{\mathrm{I}}-\frac{\mathrm{N}}{\mathrm{A}}$

In the Equation (7): y is the distance between the neutral axis and the tensile surface; I is section inertia moment (In this case I = t³/12); N is for axial force; A is compression area (A = t). Let y = μt and μ be between 0 and 0.5; then, let 12μ = a (a is a constant). From Equation (7):

(8)${\sigma }_T=\left(\frac{{\mathrm{ah}}_{\mathrm{j}+1}}{{\mathrm{t}}^2}+\frac{2-\mathrm{a}}{2\mathrm{t}}\tan \mathsf{\phi }\right)P_{j+1}-\left(\frac{2+\mathrm{a}}{2\mathrm{t}}\tan \mathsf{\phi }+\frac{{\mathrm{ah}}_{\mathrm{j}}}{{\mathrm{t}}^2}\right)P_j+\frac{\mathrm{ah}-2\mathrm{t}}{2{\mathrm{t}}^2}\mathrm{Wcos}\mathsf{\beta }-\frac{12\left({\mathrm{C}}_{\mathrm{j}+1}+{\mathrm{C}}_{\mathrm{j}}\right)\mathrm{y}}{{\mathrm{t}}^3}$

When the block only slips, its failure is mainly controlled by shear force. This shear force can be divided into two parts: slip force and anti-slip force. The resultant force of the external force on the block is denoted as F, which can be expressed as:

(9)$F=P_{j+1}+\mathrm{Wsin}\mathsf{\beta }-P_j$

The average shear stress at the bottom of a block can be expressed as:

(10)$\tau =\frac{F}{\mathrm{t}}=\left(P_{j+1}+\mathrm{Wsin}\mathsf{\beta }-P_j\right)/\mathrm{A}$

The tensile strength of block root is denoted as $\left[{\mathsf{\sigma }}_{\mathrm{t}}\right]$ and the shear strength is denoted as $\left[\mathsf{\tau }\right]$. Ideally, when ${\sigma }_T>\left[{\mathsf{\sigma }}_{\mathrm{t}}\right]$, and the shear stress is less than the shear strength, the block only occurs toppling deformation. When $\tau$= F/A >$\left[\mathsf{\tau }\right]$ and the tensile stress is less than the tensile strength, the block only slips. When a landslide occurs, shear forces and bending moments often act together on rock blocks. The different ratios of shear and bending moments in each rock block result in the different motion characteristics of each block. From the perspective of the entire landslide, it is the difference between slip deformation and toppling deformation.

4.4. Potential Causes of Slope Failure

It is difficult to give determined triggers to an ancient landslide as the area of an ancient landslide may have experienced complicated geological processes such as uplift, rainfall, and earthquakes. For anti-dip rock slopes, the most likely triggers of landslides include rock mass properties, earthquakes, and pore water [45,46,47,48].

The information about pore water pressure and possible earthquakes when the Zongrongcun landslide occurred is not clear as it is an ancient landslide. In this work, a highly idealized cantilever beam model is used to analyze the possible causes of the landslide (Figure 8).

Hoek and Bray [43] gave five secondary toppling failure types to toppling deformations of layered slopes, which included slip and toppling in toe, slip and toppling in foundation, slip and toppling in the crest of the slope, plastic flow and toppling, and pull-apart and toppling. As for the Zongrongcun landslide, it can be a little different from the five failure types above.

As shown in Figure 8, the area in the toe of the landslide is a significant slip area. The rock at the foot of the landslide slipped. The rock block near the fault belongs to the significant toppling area. However, when it comes to the headscarp of the landslide, the rock blocks there performed more slip deformation again. Therefore, we preliminarily believe that the landslide can be divided into two main damage areas: slippage and toppling. Most of the rock blocks near the fault F5b are toppling deformations, and the toe and top of the landslide have slip deformations.

As far as the entire landslide process is concerned, we can see from Figure 9 that the toe of the slope was damaged first, followed by the rock mass close to the fault F5b. Then, overall deformation occurred in the middle and headscarp of the landslide. In addition, we found traces of friction between the rocks behind the landslide during the field investigation (Figure 11a).

In general, the Zongrongcun landslide can be mainly divided into four stages (Figure 12). (1) The toe of the slope was eroded by Jinsha river throughout the year, causing the toe of the slope to be destroyed and forming a new free face. (2) After the formation of the free face, the slope deformed towards the free face because of its own weight stress and horizontal tectonic stress. This deformation triggered tensile fractures in the rock mass. (3) The tensile fracture of each rock block gradually connected to others and became a whole failure surface. (4) After the through failure surface was formed, the slope slid along the failure surface; in other words, the landslide occurred.

After the landslide occurred, because it was located at the alpine valley area and the landslide volume was huge, large numbers of landslide bodies rushed to the opposite bank and into the Jinsha River, causing river-damming. In addition, obvious lacustrine sedimentary soil can be found upstream of the landslide (Figure 11b).

4.5. Limitations and Prospects

Research in this work is aimed at replicating an ancient anti-dip river-damming landslide. Although a detailed field investigation, theoretical analysis, and numerical simulation were conducted, there are still some limitations of this research. For an ancient landslide, it is not practical to precisely determine the value of pore pressure, tectonic stress, weather, and other time-sensitive factors when the landslide occurred. That is to say, this ancient landslide may be also affected by these factors, which cannot be precisely determined. Despite all this, the simulation in this work still gives an intuitive understanding of the whole movement process of an ancient anti-dip river-damming landslide, which helps us to better reveal potential mechanisms and characteristics of this type of landslide. In terms of the prospects of this work, the concept of simulating the whole process movement for landslides can be used in many other landslide studies, which can be helpful to reveal mechanisms and characteristics for these landslides. Whole process numerical simulation can also be helpful for engineering and construction at alpine and gorge regions.

5. Conclusions

This paper reports a study of a massive ancient river-damming landslide related to toppling failure in the Tibetan Plateau. The landslide, with an estimated debris volume of 4.3 × 10⁷ m³, is located at the upper Jinsha River, SE Tibetan Plateau. It once formed a landslide dam which blocked Jinsha river. Field investigation suggests that the landslide can be divided into two zones based on geomorphic features. Each zone has its own characteristics which provide much information about the landslide mechanisms.

Theoretical analysis has been conducted to investigate the potential mechanisms and causes of the Zongrongcun landslide. The results show that the river-damming landslide was more likely to be triggered by river erosion, heavy rainfall, weight stress, and horizontal tectonic stress. These factors work together to lead to this ancient landslide.

In this work, the potential mechanisms of Zongrongcun landslide have been revealed and can be mainly divided into four stages: deformation of the slope toe stage, tensile fracture stage, formation of the failure surface, and the overall sliding. Under strong valley trenching, the rocks on the slope fractured under gravity and tectonic stress. These factors caused rock block tensile fracture failures. Then, a penetrating sliding surface formed on the slope, which subsequently caused this river-damming landslide.

Author Contributions

Funding acquisition: A.Y.; investigation, Y.G. and Y.L.; writing—original draft, Y.L.; numerical simulation, Y.L. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare that there are no conflict of interest regarding the publication of this paper.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Fault distribution in the study area (modified from Chang [41]).

Figure 2. Location and geological map of Zongrongcun landslide. (a) Location of Zongrongcun landslide. (b) Geological profile of I-I’. (c) Plan view of landslide area (c was modified from Gong et al. [42]).

Figure 2. Location and geological map of Zongrongcun landslide. (a) Location of Zongrongcun landslide. (b) Geological profile of I-I’. (c) Plan view of landslide area (c was modified from Gong et al. [42]).

Figure 3. Rocks exposed in the study area and their mineral compositions. Abbreviations: Pl—plagioclase; Qtz—quartz; Hbl—hornblende; Bt—biotite; Ms—muscovite; Kfs—potash feldspar.

Figure 4. DEM model before the landslide occurred. (a) DEM model; (b) Joint distribution in rocks.

Figure 5. Schematic diagram of river erosion.

Figure 6. Distribution of monitoring points on DEM model.

Figure 7. Front view of landslide. (a) Calculation result of Case 1; (b) Calculation result of Case 2; (c) Panorama after the landslide in Case 3.

Figure 8. Variation for Df in different position.

Figure 9. Average x-displacement for monitoring parts 1 to 5 during the landslide.

Figure 10. Simplified cantilever beam model for blocks.

Figure 11. (a) Scratch in headscarp of the landslide. (b) Lacustrine sedimentary soil upstream of the landslide.

Figure 12. The four stages of the Zongrongcun landslide. (a) Slope toe toppling damaged. (b) Occurrence of tensile failure. (c) Formation of failure surface. (d) Landslide occurred.

Table 1

Methods applied for stability on anti-dip slopes.

Methods Type	Advantage	Disadvantage
Analytical methods	Concise and easy to use	Many assumptions
Physical model tests	Effective and easy to observe the whole process	Expensive and time-consuming
Numerical simulation	Effective and not expensive	Model needs verification

Table 2

Lithology and physical and mechanical parameters (obtained from indoor geotechnical tests).

Lithology	T3δoQuartz Diorite	Pt2X3Mesoproterozoic Mica Mchist	DTJPaleozoic Ophiolite
Unit weight (kN/m3)	25.4	26.3	28.6
Uniaxial compressive strength (GPa)	1.104	0.891	0.96
Uniaxial tensile strength (GPa)	0.0123	0.0092	0.0102
Elastic modulus (GPa)	25.371	23.07	23.844
Poisson ratio	0.19	0.19	0.19
Cohesive force (MPa)	18.4	14.3	15.6
Internal friction angle	47.8°	48.9°	48.5°
Layer thickness	9~12 m/layer	3~5 m/layer	6~9 m/layer

Table 3

Structure planes in DEM model.

Name	Position	Dip Direction	Dip Angle	Quantity
Rock contact surface	Interface of mica schist and diorite	90°	88°	1
Fault f5	Interbed interface of mica schist and Plagioamphibolite	90°	70°	1
Joint group 1	Plagioamphibolite	13°	47°	Spacing 20 m
Joint group 2	Plagioamphibolite	65°	82°	Spacing 20 m
Joint group 3	Plagioamphibolite	270°	38°	Spacing 20 m
Schistosity	Mica schist	287°	68°	Spacing 20 m
Main control structural plane of headscarp	Plagioamphibolite (Slope headscarp)	264°	47°	1

Table 4

Mechanical parameters of structural planes.

Name	TangentialStiffness(GPa)	NormalStiffness(GPa)	TensileStrength(kPa)	CohesiveForce(MPa)	InternalFrictionAngle
Rockcontactsurface	0.6	0.2	20	0.26	30°
Fault f5	0.6	0.1	18	0.26	10°
Joint 1	0.1	0.1	22	0.24	28°
Joint 2	0.1	0.1	23	0.10	15°
Joint 3	0.1	0.1	25	0.05	5°
Schistosity,	0.1	0.1	26	0.05	5°
Structural plane in headscarp of landslide	0.1	0.1	16	0.01	5°

Table 5

Numerical simulation working conditions.

Steps	Case 1 (Rain Infiltration Softening, Total Reduction Factor)	Case 2 (River Undercut, Total Undercut Depth) (m)	Case 3 (Rainfall + River Cut)
1	10%	3	10% + 6 m
2	20%	6	20% + 6 m
3	30%		30% + 6 m

References

1. Froude, M.J.; Petley, D.N. Global fatal landslide occurrence from 2004 to 2016. Nat. Hazards Earth Syst. Sci.; 2018; 18, pp. 2161-2181. [DOI: https://dx.doi.org/10.5194/nhess-18-2161-2018]

2. Huang, R. Some catastrophic landslides since the twentieth century in the southwest of China. Landslides; 2009; 6, pp. 69-81. [DOI: https://dx.doi.org/10.1007/s10346-009-0142-y]

3. Xu, L.; Meng, X.; Xu, X. Natural hazard chain research in China: A review. Nat. Hazards; 2014; 70, pp. 1631-1659. [DOI: https://dx.doi.org/10.1007/s11069-013-0881-x]

4. Zhu, S.; Shi, Y.; Lu, M.; Xie, F. Dynamic mechanisms of earthquake-triggered landslides. Sci. China Earth Sci.; 2013; 56, pp. 1769-1779. [DOI: https://dx.doi.org/10.1007/s11430-013-4582-9]

5. Keefer, D.K.; Larsen, M.C. Assessing Landslide Hazards. Science; 2007; 316, pp. 1136-1138. Available online: https://www.jstor.org/stable/20036323 (accessed on 20 April 2022). [DOI: https://dx.doi.org/10.1126/science.1143308]

6. Fan, X.; Xu, Q.; van Westen, C.J.; Huang, R.; Tang, R. Characteristics and classification of landslide dams associated with the 2008 Wenchuan earthquake. Geoenvironmental Disasters; 2017; 4, pp. 1-15. [DOI: https://dx.doi.org/10.1186/s40677-017-0079-8]

7. Fan, X.; Dufresne, A.; Subramanian, S.S.; Strom, A.; Hermanns, R.; Stefanelli, C.T.; Hewitt, K.; Yunus, A.P.; Dunning, S.; Capra, L. et al. The formation and impact of landslide dams—State of the art. Earth-Sci. Rev.; 2020; 203, 103116. [DOI: https://dx.doi.org/10.1016/j.earscirev.2020.103116]

8. Zhang, Y.; Huang, C.C.; Pang, J.; Zhou, Y.; Shang, R.; Zhou, Q.; Guo, Y.; Liu, T.; Hu, G. OSL dating of the massive landslide-damming event in the Jishixia Gorge, on the upper Yellow River, NE Tibetan Plateau. Holocene; 2015; 25, pp. 745-757. [DOI: https://dx.doi.org/10.1177/0959683614567885]

9. Tian, S.; Chen, N.; Wu, H.; Yang, C.; Zhong, Z.; Rahman, M. New insights into the occurrence of the Baige landslide along the Jinsha River in Tibet. Landslides; 2020; 17, pp. 1207-1216. [DOI: https://dx.doi.org/10.1007/s10346-020-01351-4]

10. Wartman, J.; Montgomery, D.R.; Anderson, S.A.; Keaton, J.R.; Benoît, J.; Chapelle, J.D.; Gilbert, R. The 22 March 2014 Oso landslide, Washington, USA. Geomorphology; 2016; 253, pp. 275-288. [DOI: https://dx.doi.org/10.1016/j.geomorph.2015.10.022]

11. Li, Y.; Chen, J.; Zhou, F.; Song, S.; Zhang, Y.; Gu, F.; Cao, C. Identification of ancient river-blocking events and analysis of the mechanisms for the formation of landslide dams in the Suwalong section of the upper Jinsha River, SE Tibetan Plateau. Geomorphology; 2020; 368, 107351. [DOI: https://dx.doi.org/10.1016/j.geomorph.2020.107351]

12. Wu, L.Z.; Zhao, D.J.; Zhu, J.D.; Peng, J.B.; Zhou, Y. A Late Pleistocene river-damming landslide, Minjiang River, China. Landslides; 2020; 17, pp. 433-444. [DOI: https://dx.doi.org/10.1007/s10346-019-01305-5]

13. Huang, S.-Y.; Chen, Y.-G.; Burr, G.S.; Jaiswal, M.K.; Lin, Y.N.; Yin, G.; Liu, J.; Zhao, S.; Cao, Z. Late Pleistocene sedimentary history of multiple glacially dammed lake episodes along the Yarlung-Tsangpo river, southeast Tibet. Quat. Res.; 2014; 82, pp. 430-440. [DOI: https://dx.doi.org/10.1016/j.yqres.2014.06.001]

14. Dortch, J.M.; Owen, L.A.; Haneberg, W.C.; Caffee, M.W.; Dietsch, C.; Kamp, U. Nature and timing of large landslides in the Himalaya and Transhimalaya of northern India. Quat. Sci. Rev.; 2009; 28, pp. 1037-1054. [DOI: https://dx.doi.org/10.1016/j.quascirev.2008.05.002]

15. Huang, R.; Li, Y.; Yan, M. Engineering geological analysis of slope dumping deformation. J. Eng. Geol.; 2017; 25, pp. 1165-1181. [DOI: https://dx.doi.org/10.13544/j.cnki.jeg.2017.05.001] (In Chinese)

16. Huang, R. Engineering Geological Analysis of High Rock Slope Stability; Science Press: Beijing, China, 2012; (In Chinese)

17. Azarafza, M.; Akgün, H.; Ghazifard, A.; Asghari-Kaljahi, E.; Rahnamarad, J.; Derakhshani, R. Discontinuous rock slope stability analysis by limit equilibrium approaches—A review. Int. J. Digit. Earth; 2021; 14, pp. 1918-1941. [DOI: https://dx.doi.org/10.1080/17538947.2021.1988163]

18. Ashby, J. Sliding and Toppling Modes of Failure in Models and Jointed Rock Slopes; Imperial College, University of London: London, UK, 1971.

19. Alejano, L.R.; Gómez-Márquez, I.; Martínez-Alegría, R. Analysis of a complex toppling-circular slope failure. Eng. Geol.; 2010; 114, pp. 93-104. [DOI: https://dx.doi.org/10.1016/j.enggeo.2010.03.005]

20. Alejano, L.R.; Carranza-Torres, C.; Giani, G.P.; Arzúa, J. Study of the Stability Against Toppling of Rock Blocks with Rounded Edges Based on Analytical and Experimental Approaches. Eng. Geol.; 2015; 195, pp. 172-184. [DOI: https://dx.doi.org/10.1016/j.enggeo.2015.05.030]

21. Alejano, L.R.; Sánchez–Alonso, C.; Pérez–Rey, I.; Arzúa, J.; Alonso, E.; González, J.; Beltramone, L.; Ferrero, A.M. Block Toppling Stability in the Case of Rock Blocks with Rounded Edges. Eng. Geol.; 2018; 234, pp. 192-203. [DOI: https://dx.doi.org/10.1016/j.enggeo.2018.01.010]

22. Amini, M.; Majdi, A.; Aydan, Ö. Stability Analysis and the Stabilisation of Flexural Toppling Failure. Rock Mech. Rock Eng.; 2009; 42, pp. 751-782. [DOI: https://dx.doi.org/10.1007/s00603-008-0020-2]

23. Goodman, R.E. Toppling of Rock Slope, ASCE. Speciality Conf. Rock Eng. Found. Slopes; 1976; 2, pp. 201-234.

24. Amini, M.; Majdi, A.; Veshadi, M.A. Stability Analysis of Rock Slopes Against Block Flexure Toppling Failure. Rock Mech. Rock Eng.; 2012; 45, pp. 519-532. [DOI: https://dx.doi.org/10.1007/s00603-012-0220-7]

25. Azarafza, M.; Asghari-Kaljahi, E.; Ghazifard, A.; Akgün, H. Application of fuzzy expert decision-making system for rock slope block-toppling modeling and assessment: A case study. Model. Earth Syst. Environ.; 2021; 7, pp. 159-168. [DOI: https://dx.doi.org/10.1007/s40808-020-00877-9]

26. Azarafza, M.; Bonab, M.H.; Akgun, H. Numerical analysis and stability assessment of complex secondary toppling failures: A case study for the south pars special zone. Geomech. Eng.; 2021; 27, pp. 479-493. [DOI: https://dx.doi.org/10.12989/gae.2021.27.5.481]

27. Lv, Y.Q.; Liu, H.F.; Yang, J.J. Study of Failure Mechanism of Anti-Dip Rocks Materials Slope under Mining Subsidence. Adv. Mater. Res.; 2010; 143-144, pp. 1097-1101. [DOI: https://dx.doi.org/10.4028/www.scientific.net/AMR.143-144.1097]

28. Dong, M.; Zhang, F.; Lv, J.; Fei, Y.; Li, Z. Study of Stability Influencing Factors of Excavated Anti-Dip Rock Slope. KSCE J. Civ. Eng.; 2020; 24, pp. 2293-2303. [DOI: https://dx.doi.org/10.1007/s12205-020-1412-4]

29. Zhang, X.; Wang, H.; Tao, Z.; Zhu, C. Research on the Controlling Effect of NPR Cables for Anti-Dip Slope Based on the Numerical Simulation. Shock Vib.; 2021; 2021, pp. 1-11. [DOI: https://dx.doi.org/10.1155/2021/1370398]

30. Zhigang, T.; Chun, Z.; Manchao, H.; Kuiming, L. Research on the safe mining depth of anti-dip bedding slope in Changshanhao Mine. Géoméch. Geophys. Geo-Energy Geo-Resour.; 2020; 6, pp. 1-20. [DOI: https://dx.doi.org/10.1007/s40948-020-00159-9]

31. Azarafza, M.; Asghari-Kaljahi, E.; Akgün, H. Assessment of discontinuous rock slope stability with block theory and numerical modeling: A case study for the South Pars Gas Complex, Assalouyeh, Iran. Environ. Earth Sci.; 2017; 76, pp. 1-15. [DOI: https://dx.doi.org/10.1007/s12665-017-6711-9]

32. Azarafza, M.; Akgün, H.; Ghazifard, A.; Asghari-Kaljahi, E. Key-block based analytical stability method for discontinuous rock slope subjected to toppling failure. Comput. Geotech.; 2020; 124, 103620. [DOI: https://dx.doi.org/10.1016/j.compgeo.2020.103620]

33. Xie, L.; Yan, E.; Wang, J.; Lu, G.; Yu, G. Study on evolutionary characteristics of toppling deformation of reverse-dip layered rock slope based on surface displacement monitoring data. Environ. Earth Sci.; 2018; 77, 156. [DOI: https://dx.doi.org/10.1007/s12665-018-7352-3]

34. Zhang, S.-L.; Lv, P.-F.; Yang, X.-G.; Chen, X.-Z.; Zhou, J.-W. Spatiotemporal distribution and failure mechanism analyses of reservoir landslides in the Dagangshan reservoir, south-west China. Geomat. Nat. Hazards Risk; 2018; 9, pp. 791-815. [DOI: https://dx.doi.org/10.1080/19475705.2018.1480536]

35. Sardana, S.; Verma, A.K.; Verma, R.; Singh, T.N. Rock slope stability along road cut of Kulikawn to Saikhamakawn of Aizawl, Mizoram, India. Nat. Hazards; 2019; 99, pp. 753-767. [DOI: https://dx.doi.org/10.1007/s11069-019-03772-4]

36. Tu, G.; Deng, H.; Shang, Q.; Zhang, Y.; Luo, X. Deep-Seated Large-Scale Toppling Failure: A Case Study of the Lancang Slope in Southwest China. Rock Mech. Rock Eng.; 2020; 53, pp. 3417-3432. [DOI: https://dx.doi.org/10.1007/s00603-020-02132-0]

37. Hoek, E.; Bray, J. Rock Slope Engineering; 2nd ed. Publication of Institution of Mining & Metallurgy: London, UK, 1977.

38. Mohtarami, E.; Jafari, A.; Amini, M. Stability analysis of slopes against combined circular–toppling failure. Int. J. Rock Mech. Min. Sci.; 2014; 67, pp. 43-56. [DOI: https://dx.doi.org/10.1016/j.ijrmms.2013.12.020]

39. Amini, M.; Ardestani, A.; Khosravi, M.H. Stability analysis of slide-toe-toppling failure. Eng. Geol.; 2017; 228, pp. 82-96. [DOI: https://dx.doi.org/10.1016/j.enggeo.2017.07.008]

40. Amini, M.; Sarfaraz, H.; Esmaeili, K. Stability analysis of slopes with a potential of slide-head-toppling failure. Int. J. Rock Mech. Min. Sci.; 2018; 112, pp. 108-121. [DOI: https://dx.doi.org/10.1016/j.ijrmms.2018.09.008]

41. Chang, Z. Seismological and geological background of the 2013 Benzilan M5.9 earthquake in Yunnan. Seismol. Geol.; 2015; 37, pp. 192-207. [DOI: https://dx.doi.org/10.3969/j.issn.0253-4967.2015.01.015] (In Chinese)

42. Gong, Y.; Yao, A.; Li, Y.; Li, Y.; Tian, T. Classification and distribution of large-scale high-position landslides in southeastern edge of the Qinghai–Tibet Plateau, China. Environ. Earth Sci.; 2022; 81, pp. 1-19. [DOI: https://dx.doi.org/10.1007/s12665-022-10433-6]

43. Wang, H.; Liu, S.; Xu, W.; Yan, L.; Qu, X.; Xie, W.-C. Numerical investigation on the sliding process and deposit feature of an earthquake-induced landslide: A case study. Landslides; 2020; 17, pp. 2671-2682. [DOI: https://dx.doi.org/10.1007/s10346-020-01446-y]

44. Hoek, E.; Bray, J. Rock Slope Engineering; The Institute of Mining and Metallurgy: London, UK, 1981.

45. Adhikary, D.P.; Dyskin, A.V.; Jewell, R.J.; Stewart, D.P. A study of the mechanism of flexural toppling failure of rock slopes. Rock Mech. Rock Eng.; 1997; 30, pp. 75-93. [DOI: https://dx.doi.org/10.1007/BF01020126]

46. Brideau, M.-A.; Stead, D. Controls on Block Toppling Using a Three-Dimensional Distinct Element Approach. Rock Mech. Rock Eng.; 2009; 43, pp. 241-260. [DOI: https://dx.doi.org/10.1007/s00603-009-0052-2]

47. Yang, Z.; Liu, S.; Liu, Y.; He, C.; Yang, W. Dynamic stability analysis of bedding and anti-dipping rock slopes under repeated microseisms. Chin. J. Geotech. Eng.; 2018; 40, pp. 1277-1286. [DOI: https://dx.doi.org/10.11779/CJGE201807014] (In Chinese)

48. Cavallaro, A.; Abate, G.; Ferraro, A.; Giannone, A.; Grasso, S. Seismic slope stability analysis of rainfall-induced landslides in Sicily (Italy). Proceedings of the 7th International Conference on Earthquake Geotechnical Engineering; Rome, Italy, 22 October 2019; pp. 1672-1680.