Slow-moving landslides: kinematic analysis and movement evolution modeling

Three of the most well-known historical landslide areas in Western Greece were examined, concerning their temporal and kinematic evolution. Long-term ground displacements based on inclinometer data were analyzed. The aim of this study is to estimate the velocity trend type of movement through the years. The total time frame of kinematic evolution regarding the three currently studied landslide locations, was over 20 years for Panagopoula (1995-2017), 10 years for Karya (2005-2017), and 6 years for Platanos (2011-2017). All three landslides occur in soil or soft rock of Flysch and Neogene formations. In addition, two statistical approaches, time series decomposition, and autoregressive integrated moving average (ARIMA) method, were applied to identify the specific kinematic features as well as model movement evolution. In general, these techniques are used to improve forecasts as to better understand the time series characteristics. As a result, the movement evolution over time is substituted into a model that explains the behavior of the displacements time series. The models extracted from this analysis could overcome a common problem in long-term monitoring regarding measurement missing data due to technical matters or non-recorded periods.