Efficient Structure-Preserving Methods for Hyperbolic PDE Systems and modelling of Complex flows in Geophysics and Networks (HYPERFLOW)

This project develops new mathematical models based on non-linear balance laws and advanced numerical methods to simulate complex geophysical and biological flows.
Its core objectives include:
Advanced Numerical Schemes
Developing high-order, well-balanced implicit and semi-implicit methods for non-linear balance laws, alongside Machine Learning techniques (such as wPINNs) for steady-state approximations and reduced numerical diffusion.
Geophysical Flow Modelling
Formulating new multilayer, non-hydrostatic models for tsunamis; granular and debris flow models accounting for friction and dilatancy; and volcanic models capturing conduit magma dynamics, degassing, and temperature-dependent rheology.
Flows in Networks
Enhancing a multiscale model for flows in pipes and connected networks.
HPC & Real-Time Application
Deploying these solvers using HPC architectures to deliver fast, accurate simulations for real-world hazard assessment and early warning scenarios.
Project details



