Time-dependent evolution of rock slopes by a multi-modelling approach

This paper presents a multi-modelling approach that incorporates contributions from morpho-evolutionary modelling, detailed engineering-geological modelling and time-dependent stress-strain numerical modelling to analyse the rheological evolution of a river valley slope over approximately 102 kyr. The slope is located in a transient, tectonically active landscape in southwestern Tyrrhenian Calabria (Italy),where gravitational processes drive failures in rock slopes. Constraints on the valley profile development were provided by a morphoevolutionary model based on the correlation of marine and river strath terraces. Rock mass classes were identified through geomechanical parameters that were derived fromengineering-geological surveys and outputs of a multi-sensor slope monitoring system. The rock mass classes were associated to lithotechnical units to obtain a high-resolution engineering-geological model along a cross section of the valley. Time-dependent stress-strain numericalmodelling reproduced the mainmorpho-evolutionary stages of the valley slopes. The findings demonstrate that a complex combination of eustatism, uplift and Mass Rock Creep (MRC) deformations can lead to firsttime failures of rock slopes when unstable conditions are encountered up to the generation of stress-controlled shear zones. The multi-modelling approach enabled us to determine that such complex combinations may have been sufficient for the first-time failure of the S. Giovanni slope at approximately 140 ka (MIS 7), even without invoking any trigger. Conversely, further reactivations of the landslide must be related to triggers such as earthquakes, rainfall and anthropogenic activities. This failure involved a portion of the slope where a plasticity zone resulted from mass rock creep that evolved with a maximumstrain rate of 40% per thousand years, after the formation of a river strath terrace. This study demonstrates that the multi-modelling approach presented herein is a useful tool for estimating the progressive development of slope failures because it can highlight time-dependent continuous deformations as the major processes that drive rocky slopes to failure. This type of approach can be devoted to the best selection of risk mitigation strategies with respect to both human life and anthropic infrastructure.