Controlling factors for post-failure evolution of subaqueous landslides

Abstract

Subaqueous slope failures can evolve in various patterns. Whereas some slides run out and further evolve into highly mobile turbidity currents, others maintain a continuum-like structure and remain frontally confined. The different failure modes impose different hazard scenarios, making an understanding of their origin crucial for reliable hazard assessments. Yet many of the factors controlling their post-failure behaviour remain not well understood. In this paper, a procedure is presented for the analysis of the post-failure evolution of subaqueous landslides, using a coupled Eulerian–Lagrangian finite-element analysis approach. The procedure is applied to analyse the St Niklausen slide in Lake Lucerne (Switzerland), providing a good prediction of observed landslide features, without back-calibration of input parameters. The approach is further applied to investigate the influence of controlling parameters, highlighting effects of different strength parameters, the surrounding water and the slope stratigraphy and geometry on the landslide dynamics. The procedure presented can be generally applied to predict the post-failure evolution of subaqueous landslides and facilitates the extraction of the moving soil–water boundary, which can serve as source input for tsunami propagation modelling.

Depth integrated modelling of submarine landslide evolution

“…Criteria for catastrophic shear band propagation and slab failure with specific geometry and loading conditions have been recently developed and applied into back analyses of several historical events (Stoecklin 2019;Zhang et al 2020). In addition, large deformation numerical modellings of progressive failure have extended the SBP analysis from shear band propagation to complex post-failure behaviours (Dey et al 2016;Zhang et al 2019;Stoecklin et al 2020). However, these large deformation numerical methods largely rely on heavy programmes and expert users in addition to the low computational efficiency, which hence limit their practical applications.…”

“…The travel distance of the failed mass (from the main scar to the mass transport deposit limit) is up to 1500 m and the width of main scar is about 1200 m (Schnellmann et al 2002(Schnellmann et al , 2005. Such scale of subaqueous failure event was considered to trigger a tsunami hazard with the run-up height up to 4 m, and thus has received considerable attention (Schnellmann et al 2002(Schnellmann et al , 2005Stegmann et al 2007;Strasser , 2011, including a comprehensive large deformation numerical analysis using the Coupled Eulerian-Lagrangian (CEL) method (Stoecklin et al 2020). The post-failure morphology of the event has been documented using a high-resolution seismic system (Schnellmann et al 2005) and the soil properties of the intact and failed mass have been investigated through both in situ and lab geotechnical testings (Strasser et al 2007).…”

“…16a. Figure 16 b provides the initial geometry of the sectional model used in the numerical analysis, where the pre-failure lake bottom was reconstructed assuming the initial surface was parallel to the identified slide surface (Strasser et al 2011;Stoecklin et al 2020). The peak undrained strength profile of the site near the scar was obtained by Strasser et al (2007) as replotted in Fig.…”

“…17 Friction drag coefficient, C f 0.005 value in the sliding layer and weak layer being 6.1 kPa and 6.0 kPa, respectively. Soil sensitivity, which is the ratio between the peak and residual undrained shear strengths, is between 1.9 and 2.1 (Stoecklin et al 2020) with the averaged value 2.0 used in present study. The shear displacement required for the sensitive soils fall to the steady state typically ranges between 0.1 and 0.5 m (Skempton 1985) and was set to 0.15 m in this analysis.…”

“…The pre-failure lateral earth pressure coefficient was set to 0.6, considering normally consolidated conditions. The initial failure was assumed to be triggered by an earthquake of horizontal pseudostatic acceleration of 0.2 m/s 2 lasting for 2 s, which is consistent with the setting in Stoecklin et al (2020). Considering the sliding mass length of~500 m and its surface roughness of 0.01-0.1 m, the skin friction drag coefficient is averaged at around 0.005 according to Eq.…”