Analysis of the inception

Leader: Prof. Giovanni Crosta (Università di Milano Bicocca)

Why? Landslides of the flow-type can originate because of a variety of hydro-mechanical processes, such as strength degradation, pore pressure growth and groundwater level rise, among others. At the same time, the spatial variability of ground conditions and rainfalls require the use of probabilistic tools accounting for the uncertainty of the above mentioned hydro-mechanical processes. For this reason, an integrated framework for landslide susceptibility that incorporates geomechanics with stochastic techniques is fundamental for a reliable landslide hazard assessment.

What? Multi-scale computational platforms to map spatially changes in landslide susceptibility during a rainstorm. A constitutive law capturing the susceptibility to liquefaction in loose soils was used. Coupled infiltration-deformation models have been developed in conjunction with Monte Carlo analyses to compute the spatially heterogeneous probability of failure across the study areas.

Flowchart of model computations. Monte Carlo simulations are run in parallel and the results are aggregated to produce maps of probability of failure at different output times. RF = random field, z = vertical coordinate, s = suction, FS = Factor of Safety, tf = time to failure, pf = cumulative probability of failure. (Lizárraga and Buscarnera, 2019, submitted).

How? The site of interest is discretized into slope units while spatial datasets are treated as georeferenced grids that share a common reference system, Model simulations rely on a vectorized finite element solver that performs simultaneous computations of multi-phase fluid flow and slope stability. The framework is versatile, in that it allows to model the role of the landscape heterogeneity through the definition of layered deposits and the use of random fields accounting for the lateral variability.

Preliminary results
        Back-analysis of the hydrologic response of instrumented sites in Lombardia based on laboratory measurements of water retention characteristics
        (Li et al., 2019, in preparation)

        Mapping the spatiotemporal fluctuation of the margin of safety at multiple sites of interest on the basis of historical rainfall data and synthetic rainstorm scenarios

Ongoing Research

  • Parameter estimation through inverse modeling
  • Incorporation of 3D landscape morphology into stability analyses
  • Enhancement of the constitutive law through new geomechanical laboratory tests
  • Infiltration modelling in steep slopes

Researchers involved in this project

Giuseppe Buscarnera, Giovanni Crosta, Paolo Frattini, Xiang Li, José Lizárraga, Giorgio Volpi