Experimental constraints on the slip response of a slow-moving landslide to rainfall driven pore pressure changes

Landslide motion spans a continuum from slow, steady creep to rapid catastrophic failure. However, the mechanisms controlling the timing, rate, and nature of sliding, the sensitivity of motion to perturbations driven by precipitation or human activity, and potential transitions from creep to catastrophic failure all remain poorly understood. The response of landslide basal shear zones to rainfall-driven changes in pore pressure and thus effective stress can be interpreted using rate and state friction, a framework that describes the constitutive behavior and sliding stability of frictional shear zones, and is widely applied to earthquake mechanics. Laboratory experiments provide direct constraints on these frictional properties, and thus hold the potential to illuminate the material properties and conditions that control basal slip. We investigate the frictional behavior of Oak Ridge earthflow, a slow-moving landslide in the Coast Ranges of central California hosted within a clay-rich mélange. We conduct a suite of direct shear experiments to characterize its frictional rheology, including both (1) the velocity dependence of friction measured from velocity step tests; and (2) frictional healing, or time-dependent restrengthening between slip events, measured via slide-hold-slide tests. Experiments are conducted across a range of normal stresses approximating the in-situ conditions of the active shear plane (0.3 – 2 MPa) and at sliding velocities that span the range of observed landslide creep (0.001 – 30 𝜇m/s).

The shear plane material exhibits uniformly velocity strengthening behavior, characterized by a positive rate parameter (a-b), indicating that friction increases with increased slip rate, and is consistent with stable sliding. The values of (a-b) from laboratory experiments ranges from 0.001 – 0.015, in agreement with values inferred from coupled field observations of slide motion and pore pressure. Our results suggest that velocity strengthening friction, combined with modulation of effective stress through pore pressure, can generate slip transients, providing a direct mechanistic link between laboratory scale behavior and field observations of landslide motion.

We also find that the clay rich materials entrained along the base of the slide exhibit little to no healing (𝛽 ≈ 0). Near zero healing implies that the slide does not restrengthen during extended periods of low water pressure during the dry California summer. In the absence of healing, slip velocity responds directly and immediately to changes in pore pressure, independent of the duration of dry periods. Taken together, velocity strengthening friction and little to no healing are consistent with the persistent creep observed in the field, where the slip rate is governed by the stress state, pore pressure, and rate dependence of friction. Notably, Oak Ridge earthflow has been active since at least the 1930’s (the date of first air photos). The laboratory derived frictional rheology provides a quantitative framework to explain the observed landslide slip response to changes in pore pressure and suggests that friction laws can be used not only to interpret past slide behavior, but potentially to predict landslide responses to future climate-driven hydrologic forcing or other external perturbations.