Kinematics of soft soil landslides based on the analysis of microseismic monitoring data

Landslide failures are a worldwide problem claiming human lives and causing catastrophic damages to infrastructure, with their impact able to reach that of war conflicts. For these reasons this thesis is going to investigate on the use of microseismic monitoring to study the kinematic behaviour of landslides and help towards the design of an optimised early or real time warning system. All findings are based in field experiments, using 1D and 3D short period seismometers, designed to understand the capabilities and limitations of microseismic monitoring in order to enhance the first and minimize the last. The first goal of this thesis is to understand accuracy of microseismic monitoring in locating weak seismic events. The effect of the seismic velocity model is studied against common seismological assumptions, while different aperture sizes of microseismic arrays are put into test for achieving optimum location results. Next, in order to allow the study of the expected landslide seismicity during a future landslide failure, a novel methodology, an up-scaled shear box, is designed to induce landslide like seismic signals. The proposed methodology can recreate soil slip events allowing control over different geotechnical conditions, such as stress levels, thus reproducing different landslide failure scenarios. The up-scaled shear box methodology is then used to test a novel engineered seismic source: glass shard piles placed inside the landslide's mass, triggered from the landslide's displacements. The material characteristics of glass shards don't change in time, e.g. due to saturation, and have a stable seismic signature during displacement. For these reasons glass shard piles can act as the common seismic source, overcoming the time consuming process of identifying landslide complex failure mechanisms. Finally, failure is induced to two landslide vertical faces under controlled conditions. Both landslides have identical characteristics, e.g. size and geology, with the only difference that in one of them a glass shard pile was incorporated. Visual observations, displacement and loading observations of the landslide faces, timed with a GPS clock, allowing for their detection on the seismic recordings and the characterisation of failure events. The collected data are found to correlate well with the experiments conducted with the up-scaled shear box methodology, allowing for their validation.Landslide failures are a worldwide problem claiming human lives and causing catastrophic damages to infrastructure, with their impact able to reach that of war conflicts. For these reasons this thesis is going to investigate on the use of microseismic monitoring to study the kinematic behaviour of landslides and help towards the design of an optimised early or real time warning system. All findings are based in field experiments, using 1D and 3D short period seismometers, designed to understand the capabilities and limitations of microseismic monitoring in order to enhance the first and minimize the last. The first goal of this thesis is to understand accuracy of microseismic monitoring in locating weak seismic events. The effect of the seismic velocity model is studied against common seismological assumptions, while different aperture sizes of microseismic arrays are put into test for achieving optimum location results. Next, in order to allow the study of the expected landslide seismicity during a future landslide failure, a novel methodology, an up-scaled shear box, is designed to induce landslide like seismic signals. The proposed methodology can recreate soil slip events allowing control over different geotechnical conditions, such as stress levels, thus reproducing different landslide failure scenarios. The up-scaled shear box methodology is then used to test a novel engineered seismic source: glass shard piles placed inside the landslide's mass, triggered from the landslide's displacements. The material characteristics of glass shards don't change in time, e.g. due to saturation, and have a stable seismic signature during displacement. For these reasons glass shard piles can act as the common seismic source, overcoming the time consuming process of identifying landslide complex failure mechanisms. Finally, failure is induced to two landslide vertical faces under controlled conditions. Both landslides have identical characteristics, e.g. size and geology, with the only difference that in one of them a glass shard pile was incorporated. Visual observations, displacement and loading observations of the landslide faces, timed with a GPS clock, allowing for their detection on the seismic recordings and the characterisation of failure events. The collected data are found to correlate well with the experiments conducted with the up-scaled shear box methodology, allowing for their validation