Examinando por Materia "Umbrales de lluvia"

Mostrando 1 - 1 de 1
  • No hay miniatura disponible
    Publicación
    Modeling the generation and propagation of rain-triggered landslides
    (Universidad EAFIT, 2025-11-28) Rodríguez Díaz, Andrés Camilo; Ruiz Restrepo, Daniel Felipe
    Slope stability is of great importance when the associated hazard implies a risk of human and economic losses. When a morphodynamic process materializes, there is a possibility that it may evolve into a debris and mudflow, especially when the mobilized mass is large and the phenomenon occurs in mountainous or steeply sloping terrain, where the soil mass may be channeled through watercourses or low-lying areas of the landform. Currently, the numerical modeling of such landslide phenomena is mainly performed using limit equilibrium or finite element methods. However, these approaches present major limitations when dealing with large deformations. For example, the finite element method is constrained when oscillations between tensile and compressive stresses are generated. In addition, the discretized elements must remain connected; otherwise, the analysis fails to converge. This research proposes to model the initiation and propagation of a rain-induced, large-deformation landslide. The modeling is structured in two stages: i) simulation of transient rainfall and infiltration using FEM (Finite Element Method), to generate conditions where convergence cannot be reached before failure; ii) export of the FEM results, mainly pore pressures and stress states for each element, to serve as initial conditions in an MPM (Material Point Method) environment, where the movement of the failed mass is modeled according to the characteristics defined for an idealized slope representing natural conditions. To achieve this objective, a review of the numerical and constitutive bases of the methods was carried out, together with a survey of previous research on the numerical modeling of large-deformation landslides using the two approaches mentioned above. Subsequently, the geometry of an idealized slope was defined to allow predictive modeling. For this purpose, rainfall data and terrain parameters were selected from existing information for the Greater Medellín area and the Aburrá Valley in Colombia. For the rainfall input, thresholds defined by the Early Warning System of Medellín and the Aburrá Valley (SIATA), a local hazard and risk monitoring agency, were used. A rainfall threshold can be defined, in simplified terms, as the boundary between rainfall events that can potentially trigger mass movements, usually combining data on antecedent and triggering rainfall (90-day and 7-day windows in the SIATA case). For this study, three particular threshold scenarios were selected, producing three different combinations of antecedent and triggering rainfall. Finally, conclusions are drawn regarding the adaptation of the two-stage modeling approach for this type of landslide. The methodology takes advantage of the computational efficiency of FEM for simulating the initiation of mass movement, and of MPM for modeling the subsequent propagation of the soil mass across the topography. The study highlights the potential benefits for risk management of using low-cost basic information (rainfall thresholds) as an alternative to traditional and more expensive monitoring instruments such as inclinometers and piezometers.