Quantification of the uncertainty in the modelling of unstable slopes displaying marked soil heterogeneity

This study presents the application of the Boolean Stochastic Generation (BoSG) approach to the case study of the Mortisa landslide, an active earth slide flow in the Dolomites (Italy). The approach relies on the stochastic generation of different soil profiles for which the interfingering of distinct layers within a prevailing matrix is randomized. With this technique, it is possible to quantify the error associated with the simplification of stratigraphic profiles in geotechnical models. This is particularly valuable in slope stability assessment because little data are usually available in comparison with the typically complex geometry of large and deep landslides. A detailed geomorphological analysis of the study area informs an interpretation of the sequence of processes which generated the stratigraphy of the Mortisa slope. It appears to be composed of interfingered layers of silty clay and gravel that originated from subsequent earth slide and debris flow events occurring since the Late Glacial. In order to reproduce the behavior of the Mortisa slope via numerical modelling and to design effective mitigation works, it is crucial to account for the influence of the gravel lenses deposited by ancient debris flow events on the dynamics of the whole landslide. A total of 1200 possible soil configurations of the central and most active section of the Mortisa landslide have been generated. The method has been used to estimate (i) the possible errors that might occur when the presence of the gravel layers is not accounted for and (ii) the variability deriving from the position of the layers. Moreover, through the analysis of the whole ensemble of configurations, it is possible to identify the areas within the slope where the presence (or absence) of gravel layers appears to have the greatest influence on the dynamics of the landslide. This information is useful to plan future investigations and to evaluate the most effective structural mitigation measures.