Regional-scale simulation of flowslide triggering in stratified deposits

Capturing the failure mechanisms responsible for static liquefaction is a challenging task for the zoning of landslide susceptibility. While geomechanical models accounting for solid-fluid coupling can identify locations prone to flowslides, their performance in regional settings is impacted by the lack of procedures to replicate accurately stratigraphic heterogeneities. To mitigate this inconvenience, this paper discusses the performance of a spatially-distributed model aimed at capturing distinct landslide triggering mechanisms in layered deposits. The model relies on the simultaneous computation of multiple factors of safety associated with the triggering of frictional slips and liquefaction-induced flowslides. In addition, the spatiotemporal variation of such indices is retrieved from a vectorized finite element platform solving pore pressure transients in stratified unsaturated soils. Such features enable the model to take into account the vertical heterogeneity of individual slope units, while dramatically reducing the computational cost of regional-scale analyses. For illustrative purposes, the model is used to analyze a series of documented shallow landslides that occurred in Campania (Italy) following heavy rainstorms. A comparison of model predictions with and without stratigraphic heterogeneity is provided, showing that the incorporation of site heterogeneity leads to a substantial improvement of the model's spatiotemporal performance, as well as of its ability to identify different modes of failure. Furthermore, the model is applied to four municipalities across the region of interest, each characterized by different representative stratigraphic settings. Despite some mismatches, often linked to the scarcity of site-specific information, the numerical results obtained on the basis of the available dataset shows that the proposed methodology is a robust and effective tool for landslide susceptibility mapping in stratified deposits.

Automated summary

This paper discusses the performance of a spatially-distributed model aimed at capturing distinct landslide triggering mechanisms in layered deposits. The model relies on simultaneous computation of multiple factors of safety, solving pore pressure transients, and considering vertical heterogeneity, while dramatically reducing computational cost. The model shows good performance in analyzing documented shallow landslides in Campania (Italy) and municipalities with different representative stratigraphic settings.