Complex Rainfall-Driven Kinematics of the Slow-Moving Pisciotta Rock-Slide (Cilento, Southern Italy)

The Pisciotta landslide is a slow-moving deep-seated rock slide in the Cilento region (southern Italy), which involves in the middle of its body a provincial road causing continuous interruptions of vehicular traffic. It also represents a danger for two railway tunnels crossing its foot. The landslide affects a complexly folded and tectonized turbidite series composed of intercalated calcarenites, marls and mudrocks. The principal objective of this study was the kinematic analysis of the landslide and its relationship with rainfall patterns. The analysis was based on monitoring of ground deformations carried out on 50 optical targets by topographical techniques. The latters were distributed in the middle-upper part of the landslide body, in order to set an early warning system for the provincial road. The monitoring activity was performed in the periods August 2005–March 2010 (1699 days) and May 2011–September 2013 (721 days). Besides the analysis of basic kinematic parameters, such as displacement, direction of movement and velocity, also the strain field was modelled and analyzed through the SSPX numerical code. Among principal results, spatial distribution of displacements, velocities and deformation patterns revealed a strong coherence with the detected surficial structures of the landslide. Moreover, the coupled spatial and temporal variability of deformation patterns allowed understanding a complex kinematics of the landslide body whose movements were related to a series of interacting blocks of flysch rock-masses. The analysis of the temporal variability of the landslide kinematics revealed uncommon findings such as that the landslide was always in motion (active) in the observation periods, with velocities ranging from the very low to low classes. Maximum velocities were observed unexpectedly in summer, during the dry period. Cross-correlation analyses performed between mean landslide velocity and rainfall time series showed a delay of about six months.