DYNAMIC EARTH-PRESSURE CHANGES ASSOCIATED WITH MOVEMENT OF A SLOW LANDSLIDE (Invited Presentation)

Long-term continuous monitoring of landslides and unstable slopes usually involves long-period (subhourly to daily) measurements of pore pressure, displacement, and related variables to characterize temporal changes in stability. Use of long-period measurements is often justifiable based on power requirements, limitations of field instrumentation for storing or transmitting data, data management, and especially, the slow rate of change of conditions in many landslides. Nevertheless, noteworthy and sometimes abrupt changes in conditions related to reactivation or slow movement of landslides may happen over very short time spans (minutes, or less) as revealed by kinematic and hydrological data collected over many years. Low-frequency monitoring either misses such changes entirely or leaves a deficient record. In an effort to observe dynamic processes in a landslide complex near Puget Sound, Washington, we used new, high-speed vibrating-wire technology to combine short-period (0.01 s) measurements of earth and pore-water pressure at two locations with 15-minute measurements of hydrologic conditions, and a time-lapse camera (subdaily). The camera recorded six separate episodes of movement during January and March 2016. The 15-minute measurements of soil water content and pore-water pressure indicate wet soil and relatively high pore pressures associated with these movement episodes. We sampled total earth pressure and pore-water pressure at 1 s for 10 minutes daily to obtain a baseline record of changes in effective earth pressure over a nine-month period of observation and used rainfall to trigger intervals of continuous 0.01-s sampling. Both short-period sensor arrays showed a delayed, gradual 7-kPa increase of effective earth pressure over many hours between the first and second movement episodes in January. Changes of 1 kPa were observed between and during movement in March. Elevated earth pressures fluctuated slightly (January – April) and then decayed to premovement pressures in late April and May. We expect details revealed by the short-period data to lead to new insights into stress changes accompanying landslide reactivation and movement.