Answers are grounded in this paper and cite their sources.
: The question of the accuracy of numerical schemes for hyperbolic systems with source terms around stationary solutions and/or in asymptotic regimes has been a subject of great interest over the last two decades, see the seminal works [2,11,12] in late nineties and the reference books [4,10] ten years later. In the context of geophysical flows and for colocated finite-volume methods applied to shallow water equations, a lot of works have been devoted to the accuracy around the so-called lake-at-rest equilibrium and more recently extended to nonzero velocity one dimensional stationary states, see [3] and references therein. But for large scale atmospheric or oceanographic flows, the relevant stationary state is the geostrophic equilibrium, see [21] for a general introduction to geophysical rotating fluid dynamics. The accuracy of colocated finite-volume numerical schemes around such an equilibrium was less investigated. To our knowledge, the first work in this field is due to Bouchut, Le Sommer and Zeitlin [5], see also [6, 7], but was fully accurate only for one-dimensional flows, as exhibited in [1]. Recently two independent works [15,20] proposed IMplicit-EXplicit type schemes for fully nonlinear equations which are proven to be accurate near the geostrophic equilibrium but, due to their implicit part, need to solve a global Laplace equation at each time step. Note that there exists also a lot of works devoted to the approximation of the Coriolis term in staggered finite-difference schemes, see for example [19] for a linear analysis and [18] for the fully nonlinear case and in the finite-element framework [16]. In
No takes yet. Share an insight, caveat, or question.
Audusse et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: