Improved understanding of slope failure in porphyry rocks by integrated rock mass characterisation, remote sensing and numerical modelling

Rock slopes fail by structurally-controlled mechanisms, global circular failures, or complex mechanisms depending on structural patterns and the type and amount of rock mass damage. When dealing with large natural or engineered rock slopes (e.g. quarry or open-pit slopes) in complex geological settings, spatial and temporal interplays among structurally-controlled and global failure modes can be difficult to assess, although they are important in a geohazard perspective, e.g. to select proper monitoring, analysis and risk reduction strategies. The mode and timing of global slope failures is controlled by rock mass structure and strength, topography, external triggers (and their evolution in time), and by processes involving progressive failure and extensive rock damage. An improved understanding of global rock slope failure can thus be achieved by integrating: 1) rock mass structure and damage characterisation by structural geology and rock mechanics, exploiting both field and remotely sensed data; 2) spatially-distributed monitoring of slope dispiacements; 3) numerical modelling linking structurally-controlled and global failure mechanisms on different spatial scales. We used the Mt. Gorsa quarry slope (Trentino, Italy) to test an analysis approach integrating rock mass characterisation, Terrestrial Laser Scanning (TLS), ground-based radar interferometry (GB-InSAR) and Finite Element modelling. Mt. Gorsa slope has been excavated in strongly anisotropic rhyolitic ignimbrites, was affected in 2003 by a major roto-translational rockslide involving about 400,000 m3 of disrupted rock, and undergoes continuing instability monitored by GB-InSAR. Structural analysis of TLS and field data shows that the slope is affected by widespread local, structurally-controlled instabilities (sliding, toppling, strain localizationin kink bands). To investigate the non-obvious relationships between structurally-controlled and global slope failure mechanisms, we quantified rock mass damage along the slope using a new approach, by mapping the Geological Strength Index (GSI) in the field and interpreting its topographic signatures in TLS point clouds. We systematically mapped a persistent geological marker in TLS point clouds, and established correlations between attitude variability statistics and rock mass damage, providing an efficient assessment tool at the slope scale. Rock mass damage increases in slope sectors affected by structurally-controlled instability and is maximum in areas of ongoing global instability. Field data and GB-InSAR monitoring show that the 2003 rockslide occurred inside a damaged rock mass zone, characterized by GSI values representing threshold conditions for global slope failure. Statistical analysis allowed establishing site-specific, empirical relationships among TLS-derived variability statistics, GSI values, and slope displacements. We used Finite-Element numerical modelling to integrate all available data, and suggest that rock mass damage induced by local, structurally-controlled slope instability provides the required conditions (loss of structural pattern, block size reduction, cohesion loss) for transition to overall equivalent continuum behaviour and global slope failure, which can be described and predicted in terms of "damage paths".