A comparative analysis of mesh-based and particle-based numerical methods for landslide run-out simulations
Keywords
High Performance Computing
Advanced Numerical Methods for Scientific Computing
Geosciences/Protection of Land and Water Resources
Code:
103/2024
Title:
A comparative analysis of mesh-based and particle-based numerical methods for landslide run-out simulations
Date:
Tuesday 3rd December 2024
Author(s):
Fois, M.; Gatti, F.; de Falco, C.; Formaggia, L.
Abstract:
Landslides are among the most dangerous natural disasters, with their unpredictability and potential for catastrophic human and economic losses exacerbated by climate change. Continuous monitoring and precise modeling of landslide-prone areas are crucial for effective risk management and mitigation. This study explores two distinct numerical simulation approaches: the mesh-based finite element model and the particle-based model. Both methods are analyzed for their ability to simulate landslide dynamics, focusing on their respective advantages in handling complex terrain, material interactions, and large deformations. A modified version of the second-order Taylor-Galerkin scheme and the depth-averaged Material Point Method are employed to model gravity-driven free surface flows, based on depth-integrated incompressible Navier-Stokes equations. The methods are rigorously tested against benchmarks and applied to a real-world scenario to assess their performance, strengths, and limitations. The results offer insights into selecting appropriate simulation techniques for landslide analysis, depending on specific modeling requirements and computational resources.