Avraam Konstantinidis

Abstract:

A statistical study of precursor activity in rain- as well as earthquake-induced landslides by means of spring block models enhanced with displacement gradients and stochasticity is performed. This way, a robust 2D model can be formulated for studying triggered landslides. A cellular automaton is utilized, in order to examine the dynamic behavior and the stability of rock/soil slopes due to neighboring heavy rainfall or earthquake activity. The type and nature of the failure plane, as well as the triggering mechanism is studied. Moreover, the different dynamic evolution modes of the slope can be mapped to specific shape parameters of the corresponding distributions of the incremental displacements. These parameters are calibrated through comparison with statistical data on landslides events available in the literature. The calibrated model can then be used as a means for understanding, predicting and mitigating the impact of catastrophic landslides and its theoretical predictions are compared with the respective predictions of alternative techniques for studying slope stability.