Geophysical-geotechnical sensor networks for landslide monitoring

Landslides are often the result of complex, multi-phase processes where gradual deterioration of shear strength within the sub-surface precedes the appearance of surface features and slope failure. Moisture content increases and the build-up of associated pore water pressures are invariably associated with a loss of strength, and thus are a precursor to failure. Consequently, hydraulic processes typically play a major role in the development of landslides. Geoelectrical techniques, such as resistivity and self-potential are being increasingly applied to study landslide structure and the hydraulics of landslide processes. The great strengths of these techniques are that they provide spatial or volumetric information at the site scale, which, when calibrated with appropriate geotechnical and hydrogeological data, can be used to characterise lithological variability and monitor hydraulic changes in the subsurface. In this study we describe the development of an automated time-lapse electrical resistivity tomography (ALERT) and geotechnical monitoring system on an active inland landslide near Malton, North Yorkshire, UK. The overarching objective of the research is to develop a 4D landslide monitoring system that can characterise the subsurface structure of the landslide, and reveal the hydraulic precursors to movement. The site is a particularly import research facility as it is representative of many lowland UK situations in which weak mudrocks have failed on valley sides. Significant research efforts have already been expended at the site, and a number of baseline data sets have been collected, including ground and airborne LIDAR, geomorphologic and geological maps, and geophysical models. The monitoring network comprises an ALERT monitoring station connected to a 3D monitoring electrode array installed across an area of 5,500 m2, extending from above the back scarp to beyond the toe of the landslide. The ALERT instrument uses wireless telemetry (in this case GPRS) to communicate with an office based server, which runs control software and a database management system. The control software is used to schedule data acquisition, whilst the database management system stores, processes and inverts the remotely streamed ERT data. Once installed and configured, the system operates autonomously without manual intervention. Modifications to the ALERT system at this site have included the addition of environmental and geotechnical sensors to monitor rainfall, ground movement, ground and air temperature, and pore pressure changes within the landslide. The system is housed in a weatherproof enclosure and is powered by batteries charged by a wind turbine & solar panels. 3D ERT images generated from the landslide have been calibrated against resistivity information derived from laboratory testing of borehole core recovered from the landslide. The calibrated images revealed key aspects of the 3D landslide structure, including the lateral extent of slipped material and zones of depletion and accumulation; the surface of separation and the thickness of individual earth flow lobes; and the dipping in situ geological boundary between the bedrock formations. Time-lapse analysis of resistivity signatures has revealed artefacts within the images that are diagnostic of electrode movement. Analytical models have been developed to simulate the observed artefacts, from which predictions of electrode movement have been derived. This information has been used to correct the ERT data sets, and has provided a means of using ERT to monitor landslide movement across the entire ALERT imaging area. Initial assessment of seasonal changes in the resistivity signature has indicated that the system is sensitive to moisture content changes in the body of the landslide, thereby providing a basis for further development of the system with the aim of monitoring hydraulic precursors to failure.