Long-term evolution and early warning strategies for complex rockslides by real-time monitoring

The potential of long-term, real-time surface displacement monitoring by ground-based radar interferometry (GB-InSAR) to improve the understanding of mechanisms and set up objective early warning criteria for complex rockslides is explored. Monitoring data for a rockslide in the Central Italian Alps, collected since 1997 by ground-based and remote-sensing techniques, are examined. A unique 9-year continuous GB-InSAR monitoring activity supported an objective subdivision of the rockslide into early warning domains with homogeneous involved material, mechanisms and sensitivity to rainfall inputs. Distributed GB-InSAR data allowed setting up a virtual monitoring network by a posteriori selection of critical locations representative of early warning domains, for which we analysed relationships among rainfall descriptors and displacement rates. The potential of different early warning criteria, depending on the instability mechanisms dominating different domains, is tested. Results show that (a) rainfall intensity-duration-displacement rate relationships can be useful tools to predict displacements of rainfall-sensitive rockslide sectors, where clear trigger-response signals occur, but are unsuitable in rockslide domains affected by the long-term progressive failure of the rock slope and (b) effective early warning strategies for collapse scenarios (entire rockslide, specific domains) can be enforced by modelling real-time, high-frequency GB-InSAR data according to the accelerated creep theory.