Flowslide Triggering in Volcanic Soils: Role of Stratigraphy and Bedrock Exfiltration

Rainfall-induced landslides pose significant risks to urban communities and infrastructure. The combination of transient rainfall, unsaturated soil properties, and spatially varying topography makes them difficult to forecast over large areas. This contribution describes a computational framework to evaluate landslide susceptibility over regional landscapes. To this aim, a description of a case study is presented, in which more than 40 shallow flowslides were triggered after 48 h of continuous precipitation over a region of 9 km². Several hypotheses have been proposed to explain the widespread distribution of slope instabilities, such as soil liquefaction, pore-pressure pulses due to layering, bedrock exfiltration, and antecedent hydrologic conditions, among others. Here, using available field and laboratory data to constrain the input parameters, different model scenarios are tested to back analyze the spatial and temporal occurrence of the events, namely, (1) slope failure caused by infiltration in homogenous deposits, (2) failures mediated by permeability contrasts in heterogeneous slopes, and (3) slope instabilities caused by bedrock springs within homogeneous soil profiles. Each scenario is evaluated by comparing the computed susceptibility maps and the temporal evolution of unstable areas against the landslide inventory. It is shown that stratigraphy effects can capture successfully the observed distribution of landslide source areas. Finally, the advantages and limitations of each scenario are discussed, and recommendations for future analyses are proposed.