Laboratory simulations of slow-moving landslide mechanisms

Abstract

Slow-moving landslides are a significant hazard in New Zealand and globally and better understanding of how they may move during and after earthquakes and rainstorms is needed. We collected samples from the shear zone of the Utiku Landslide, a slow-moving landslide complex in New Zealand. We conducted a series of experiments on the samples to simulate different pore-water pressure and dynamic stress scenarios on the samples. We compared the laboratory results with high resolution landslide monitoring data which included rainstorms and used well constrained numerical modelling of potential ground displacement during earthquakes. This allowed us to demonstrate how pore-water pressure changes and earthquake shaking might influence rates of movement in slow-moving landslides.

Our experiments showed that during periods of elevated pore-water pressure, displacement rates are influenced by pore-water pressure and the rate of change in pore-water pressure within the shear zone. During dynamic tests we showed displacement rates were controlled by the forces operating at the shear zone when they exceeded the yield acceleration. During strong earthquake accelerations, although strain should increase rapidly, this will only cause relatively minor increases in the out of balance forces.

The behaviour we observed in the laboratory was consistent with the measured landslide displacement response during fluctuations in pore-water pressures and numerical simulations of landslide displacement during earthquakes. Although large landslide displacements could occur during strong earthquake shaking, this does not occur frequently at the Utiku Landslide, or other similar landslides in the district. Consequently, landslide displacement is predominantly controlled by pore-water pressures over shorter timescales (decades) relative to seismic cycles (centuries) within the study area.

Our results showed how the mechanisms of shear-zone deformation control the movement patterns of many large, slow moving translational landslides and how they may be mobilised by strong earthquakes and rainfall.

---

Technical Abstract

Slow-moving landslides are a significant hazard in New Zealand and globally and better characterisation of their potential movement mechanisms during earthquake and rainstorm is needed. Understanding their behaviour requires a combination of high-resolution field monitoring, well-constrained ground models and laboratory testing that accurately replicates the natural stress conditions, but this combination is rarely available.

In our study we used a dynamic back-pressured shear box to simulate representative stress conditions in the slow-moving Utiku Landslide during phases of pore-water pressure fluctuation and dynamic shear. We combine this novel dataset with monitoring records and numerical modelling from Utiku landslide to provide new insight into its movement mechanisms.

Our laboratory results show that during periods of elevated pore-water pressure, displacement rates were influenced by two components: 1) an absolute stress state component (normal effective stress); and 2) a transient stress state component (the rate of change of normal effective stress). The behaviour observed in the laboratory was consistent with the ground-monitoring records and explains why the relationship between displacement rate and pore-water pressure is different during periods of acceleration and deceleration in some slow-moving landslides.

During dynamic shear experiments we showed that displacement rates were controlled by the extent to which the forces operating at the shear surface were out of balance. Once these forces exceeded the yield acceleration, displacement rates increased rapidly with distance normal to the failure envelope in plots of shear stress against normal effective stress. By combining laboratory and numerical simulation data we demonstrated that, during strong earthquake accelerations, strain should increase rapidly with relatively minor increases in the out of balance forces (reducing the KY / Kmax ratio). Therefore, we reasoned that large landslide displacements could occur when accelerated by strong earthquakes, but such accelerations in the study area occur infrequently. Thus, in this area over long timescales (i.e. multiple seismic cycles) landslide displacement will be predominantly controlled by pore-water pressures.

By combining the specialised laboratory testing with field monitoring, well-constrained ground models, and numerical simulations, we showed how the mechanisms of deformation occurring along a landslide shear surface could control the movement patterns seen in many large, slow moving translational landslides. The combining of these approaches provides a robust framework to use in the hazard assessment of landslides that could be mobilised by both strong earthquakes and significant rainfall.