Elsevier

Engineering Geology

Volume 278, 5 December 2020, 105819
Engineering Geology

Dynamics and emplacement mechanisms of the successive Baige landslides on the Upper Reaches of the Jinsha River, China

https://doi.org/10.1016/j.enggeo.2020.105819Get rights and content

Highlights

  • The stratigraphic preservation, diapiric and convoluted structures are observed in rock avalanche deposits.

  • A simple differential shear is proposed as the main emplacement form.

  • Discrete element simulations support the landslide process reconstructions

  • Overloading-impact-induced erosion is proposed to be the most significant contribution to entrainment.

Abstract

Two successive Baige landslides are presented in this study to investigate the dynamics and emplacement mechanisms. Field survey shows that the deposit characteristics of the first landslide are affected by the extremely fractured rock masses and the complex spatial distribution of lithologies in the source area. This deposit consists of abundant angular megablocks on the surface supported by clasts, with a relatively abundant content of rubble ranging from micrometre to decimetre sizes, and rarely well-preserved original stratigraphic sequence, diapiric structure, and convoluted structure. These sedimentological characteristics imply that a simple differential shear process, induced by compression, is distributed through the entire avalanche masses with low internal disturbance during the emplacement of the first landslide. The three-dimensional discrete element simulations show that: (i) the first landslide reaches its maximum average velocity of 41 m/s in ~28 s and eventually is bulldozed to a halt in the river within ~80 s. (ii) The second landslide is controlled by entrainment, which impedes the avalanche from further accelerating, and simultaneously imparts kinematic energy to the substrate, eventually becoming entirely mobilized and reaching a maximum average velocity of 35 m/s in ~36 s. In the dynamic process of the second landslide, the significant entrainment-induced enlarged volume (its entrainment ratio reaches 3.3) is associated with the steep concave terrain, water-bearing substrate and lower-roughness substrate basal boundary, and overloading-impact-induced erosion is proposed to be the most significant contribution to this entrainment.

Introduction

Rock avalanches, a kind of flow-like movement of fragmented rock masses, commonly present huge volume, extremely high velocity and high mobility (i.e., low Fahrböschung often ranging from 0.1 to 0.3) always induced by massive and rapid rockslides or rockfalls (Heim, 1932; Sassa et al., 2004; Hungr, 2006; Davies and McSaveney, 2009; Pudasaini and Miller, 2013; Wang et al., 2013; Hungr et al., 2014; Yin et al., 2016; Dufresne and Dunning, 2017). Such mass movements can dramatically impact or bury areas along transport paths, consequently leading to significant hazard to human life and infrastructure, as, for example, the 2000 Yigong landslide (Yin, 2011) and 2013 Sanxicun landslide in China (Yin et al., 2016), the 2006 Leyte landslide (Sassa et al., 2010) in Philippines, and the Flims landslide in the Alps (Pollet et al., 2005). These hazards remind us that further researches into rock avalanches are of both importance and urgency to assist in developing possible mitigating strategies (Dufresne, 2012; Zhou et al., 2016; Wang et al., 2019b).

Over the past few decades, attempts have been made to better understand how rock avalanches move, and further to provide insight into their hypermobility and flow-like motion. However, anisotropic dynamic processes and multiple overlapping factors (Dufresne and Dunning, 2017) result in contradictory hypotheses being proposed, as summarized in Hungr (2006), Zhang and Yin (2013), and Dufresne and Dunning (2017). The transport process of a landslide is generally constrained by complex mechanisms, some of which are necessary but not sufficient, or neither necessary nor sufficient, due to lack of the complete explanations of dynamic phenomena and deposit characteristics (e.g., fragmentation) (Davies and McSaveney, 2009; Dufresne and Dunning, 2017). For example, Yin and Xing (2012) employed wind tunnel modelling to show that air-cushioning was a factor aiding lubrication in the Yigong landslide. Zhao and Crosta (2018) adopted the acoustic fluidization theory to investigate the Tangjiashan landslide and indicated that the basal stress ratio played a crucial role in its transport. However, both of these mechanisms are unable to fully explain the observed deposit structures (Hungr, 2006; Dufresne and Dunning, 2017).

The dynamic fragmentation and lubrication by liquefied saturation soil hypotheses are two common focuses within existing researches, as pervasive fragmentation and displacement-liquefaction saturation substrates exist in or around landslides (Sassa et al., 2004; Hungr, 2006; Davies and McSaveney, 2009; Dufresne and Dunning, 2017). The hypotheses proposed by Davies and McSaveney (2009) indicated that once the elastic strain within the force chain network accumulated to a certain extent, fragmentation would occur and then result in an effective decrease in frictional resistance. Nevertheless, the fragmentation process is considered to be the dissipation of energy (Crosta et al., 2007), and merely results in the distal avalanche travelling further (Bowman et al., 2012) while reducing the displacement of the centre (Haug et al., 2016). Typical facies models have been proposed to describe the detailed sedimentological characteristics of deposits and have been further used to reveal the emplacement mechanisms (Dufresne et al., 2016; Dufresne and Dunning, 2017; Wang et al., 2019b). Combining field investigations and numerical simulations, Zhang et al. (2019a) studied the deposit structures and emplacement mechanisms of the Jiweishan landslide. Researches have also revealed that liquefied substrate, such as that produced via impact-induced (Sassa et al., 2004) and seismic loading-induced liquefaction (Wang et al., 2013), played a dominant role in landslide motion when the liquefied substrate is entrained and incorporated into the basal portion of the moving masses (Hungr and Evans, 2004). However, not all rock avalanche runout surfaces are filled with saturated substrate (Davies and McSaveney, 2009), and the ability of the substrate to enhance mobility also depends on the basal conditions and substrate properties (Dufresne, 2012). Thus, substrate originally underlying the transport path does not always promote the mobility of rock avalanches (Zhang et al., 2019a).

Despite extensive researches into catastrophic rock avalanches, studies on the dynamics and emplacement mechanisms remain insufficient. In this study, we focus on the Baige landslide, a typical rock avalanche occurred in the Jinsha River tectonic suture zone belonging to the upper reaches of the Jinsha River. This landslide successively experienced two large-scale mass movement processes on October 10 and November 3, 2018, respectively, causing significant economic losses due to landslide dam breaching floods (Fan et al., 2019; Zhang et al., 2020). These two landslides involve a series of spectacular dynamic processes, including high-position rockslide (i.e., initiated from ridge top with larger vertical distance, e.g., Xinmo landslide (Yin et al., 2017)), fragmentation into rock avalanche and entrainment, which provide an opportunity to study their dynamics and emplacement mechanisms by well-supported data.

Section snippets

Method

Field survey was conducted following the first landslide. In particular, further comprehensive field surveys were carried out after the second landslide through core logging, drill hole logging, and trial trench digging in order to confirm the geomorphological, stratigraphic, lithological, and tectonic characteristics. In addition, remote sensing images prior to the first landslide event (dated 28 February 2018), after the first landslide event (dated 16 October 2018), and after the second

Geological setting

The Jinsha River tectonic suture zone, situated on the eastern portion of the Qinghai Tibetan Plateau, is a complex zone composed of several faults and tectonic blocks that have been metamorphosed and deformed in multiple periods (Fan et al., 2019). The primary tectonic trace of this zone is oriented NW (Fig. 1a). The Boluo-Muxie fault (F1), the Zhuying-Gongda fault (F2) and the Zeba-Xietang fault (F3) are considered to be the boundaries of the Jinsha River tectonic suture zone (Fig. 1a) (Zhang

Basic characteristics of the Baige landslide

This section presents findings regarding well-preserved deposit structures and characteristics of the successive Baige landslides based on the field survey and remote sensing images, which can be used to reflect the dynamics and emplacement mechanisms.

DEM-based dynamic process analysis

In the field, some valuable phenomena are preserved in and around the final deposit. Nevertheless, eradication of early key features, inaccessibility of the internal exposure, insufficient field investigation strategies, and unavailable witnesses of the actual mass movement are likely to affect our ability to identify crucial implications for reconstructing the emplacement process (Dufresne, 2012; Zhang et al., 2019a). For example, dynamic parameters of the landslides and the interactive

Implications of the emplacement mechanisms of the first landslide

Many features are observed in the deposit of the first Baige landslide, particularly stratigraphic preservation, diapiric structure and convoluted structure, which are of importance to be regarded as implication for understanding the emplacement mechanisms (Dufresne, 2009; Dufresne et al., 2016; Wang et al., 2019b; Zhang et al., 2019a). The preservation of original stratigraphic sequence has been widely reported (Dufresne et al., 2016; Zhao and Crosta, 2018; Wang et al., 2019b; Zhang et al.,

Conclusions

This paper presents a combined field survey and numerical simulation for analysing the deposit characteristics and dynamic processes of two successive Baige landslides. Particular focus is given to understand dynamics and emplacement mechanisms. Our conclusions are summarized as follows:

The first landslide can be divided into the source area, the deposit zone and the wave-influence zone. The deposit characteristics depend on extremely fractured rock masses and the complex spatial distribution

Declaration of Competing Interest

None.

Acknowledgments

We gratefully acknowledge the support of the National Key R&D Program of China (2018YFC1505404), the project from China Geological Survey (DD20190637), the National Natural Science Foundation of China (41907225). Critical reviews by the editor Prof. Wasowski and the two anonymous reviewers have greatly improved this manuscript and are gratefully acknowledged.

References (39)

  • Z.X. Zou et al.

    Kinetic characteristics of debris flows as exemplified by field investigations and discrete element simulation of the catastrophic Jiweishan rockslide, China

    Geomorphology.

    (2017)
  • E.T. Bowman et al.

    Physical models of rock avalanche spreading behaviour with dynamic fragmentation

    Can. Geotech. J.

    (2012)
  • B. Cagnoli et al.

    Grain size and flow volume effects on granular flow mobility in numerical simulations: 3-D discrete element modeling of flows of angular rock fragments

    J. Geophys. Res. Solid Earth

    (2015)
  • G.B. Crosta et al.

    Fragmentation in the Val Pola rock avalanche, Italian Alps

    J. Geophys. Res. Earth Surf.

    (2007)
  • A. Dufresne

    Influence of Runout Path Material on Rock and Debris Avalanche Mobility: Field Evidence and Analogue Modelling (PhD Thesis)

    (2009)
  • A. Dufresne

    Granular flow experiments on the interaction with stationary runout path materials and comparison to rock avalanche events

    Earth Surf. Process. Landf.

    (2012)
  • A. Dufresne et al.

    Process dependence of grain size distributions in rock avalanche deposits

    Landslides.

    (2017)
  • J. Estep et al.

    Substrate effects from force chain dynamics in dense granular flows

    J. Geophys. Res. Earth Surf.

    (2012)
  • X.M. Fan et al.

    Successive landsliding and damming of the Jinsha River in eastern Tibet, China: prime investigation, early warning, and emergency response

    Landslides.

    (2019)
  • Cited by (44)

    • Insights into some large-scale landslides in southeastern margin of Qinghai-Tibet Plateau

      2023, Journal of Rock Mechanics and Geotechnical Engineering
    • Geostructures, dynamics and risk mitigation of high-altitude and long-runout rockslides

      2023, Journal of Rock Mechanics and Geotechnical Engineering
      Citation Excerpt :

      The active mobilization of this package is sliding later by layer until the residual momentum of substrate fails to motivate. Furthermore, in the process of interaction between landslide and substrate, a deformation wave ahead of the flow under continuous impact-related erosion can improve erosion of substrate (Estep and Dufek, 2012; Zhang et al., 2020a). As shown in phase-1 and phase-2 in Fig. 28, the increments of the basal pore-fluid pressures ahead of the flow front are observed at two measuring points as a result of substrate compression, mobilization, and transport, which can be used as a proxy to reflect the effect of impact-related erosion (Iverson et al., 2011).

    • Energy transfer mechanisms in flow-like landslide processes in deep valleys

      2022, Engineering Geology
      Citation Excerpt :

      The time histories of centroid velocity for flows of different solid concentrations during the entire landslide-river interaction process are shown in Fig. 10. Comparison is also made with those obtained from other numerical results (Ouyang et al., 2019; Zhang et al., 2020). The total sliding time of the Baige landslide is about 80 s. Seismic records available in the literature also show that the seismic signal of the Baige landslide becomes weak after 80 s (Zhang et al., 2019a; An et al., 2021).

    View all citing articles on Scopus
    View full text