Abstract
This paper introduces a new family of minimally implicit Runge–Kutta (MIRK) methods for the numerical approximation of hyperbolic balance laws with stiff source terms. Our main focus is the one-dimensional shallow water system with topography and Manning friction, a model of central importance in fluid dynamics. Preserving steady states, i.e. , achieving the well-balanced property, is a fundamental requirement in order to capture small perturbations around equilibria without contamination by spurious numerical effects. While implicit–explicit (IMEX) strategies based on diagonally implicit Runge–Kutta schemes have been successfully employed in this context, their efficiency is often limited by the cost of the implicit stages. The proposed MIRK schemes combine the robustness of semi-implicit approaches with a substantially reduced computational effort. They are fully well-balanced, in the sense of preserving all stationary solutions of the shallow water equations with Manning friction, while maintaining accuracy and stability for stiff regimes. Extensive numerical experiments demonstrate that MIRK methods achieve significant speed-ups compared to classical IMEX schemes, without compromising reliability.
Funding Information
  • Ministerio de Ciencia e Innovación (PID2022-137637NB-C21)
  • Ministerio de Ciencia e Innovación (PID2022-136828NB-C43)
  • Ministerio de Ciencia e Innovación (PID2021-125485NB-C21)
  • Ministerio de Ciencia e Innovación (RED2024-153735-E)
  • Ministerio de Ciencia e Innovación (PCI2024-155061-2)
  • Deutsche Forschungsgemeinschaft (DFG-AEI 2023)
  • European Horizon Europe Staff Exchange Programme (NewFunFiCO-101086251)

This publication has 55 references indexed in Scilit: