DILATANCY, PORE-PRESSURE FEEDBACK, AND REGULATION OF LANDSLIDE (AND THRUST FAULT) MOTION

Diverse styles and rates of landslide motion may result from subtle differences in dilatancy and attendant pore-pressure responses in basal shear zones. At one extreme, landslides exhibit prolonged, stable, creeping motion regulated by dilation and negative pore-pressure feedback, and at the other extreme, they exhibit runaway acceleration enhanced by contraction and positive pore-pressure feedback. Intermediate landslide behavior can involve repeated stick-slip episodes that ultimately culminate in either runaway acceleration or long-term stabilization. A simple mechanical model generates all three types of landslide behavior as a consequence of modest differences in dilatancy, hydraulic diffusivity, and assumptions about the possible rate-dependence of basal friction. No assumptions are necessary regarding “state” dependence of friction (commonly invoked in fault-slip models to account for time-dependent healing that affects friction), because transient pore-pressure responses provide an intrinsic time scale that governs evolution of the frictional “state.” Another key difference between this landslide model and most fault-slip models involves the driving force, which is specified in terms of gravitationally imposed stress rather than tectonically imposed strain. In the landslide model, gravity-driven inertial motion of a poroelastic block is coupled to pore-pressure change in a thin basal shear zone that deforms inelastically. This pore-pressure change is described by a boundary condition that assumes Darcian flow of fluid and conservation of solid and fluid mass within the deforming shear zone. With addition of a tectonic driving force, the same model may provide a suitable framework for analyzing the dynamics of low-angle thrust faults, such as those analyzed statically by Hubbert and Rubey in their landmark paper of 1959 (Geol. Soc. Amer. Bull., v. 70, p. 115-166).