Exact moment models for conservation laws in phase space
Tileuzhan Mukhamet, Katharina Kormann
TL;DR
This work addresses the computational burden of high-dimensional kinetic conservation laws by deriving exact reduced models through Burby’s centered-moment parametrization of the distribution function. It develops exact fluid moment models of degree $k$ closed by a center evolution $\partial_t \bm v = - \bm G(\bm v) \cdot (\nabla \bm v) + \bm F(\bm v)$, with explicit degree-0,1,2 closures, and constructs phase-space particle models around moving centers $\bm Z_p$ that are exact when $\partial_t \bm Z_p = \bm A(\bm Z_p,t)$, including degree-0,1,2 realizations. A hybrid fluid–particle framework combines these ansätze to preserve exactness for the sum, enabling flexible representations of plasmas. The approach is applied to non-relativistic and relativistic Vlasov–Maxwell systems, yielding invariant-preserving, computationally tractable models and addressing practical issues such as current regularization in distributional settings.
Abstract
Moment equations offer a compelling alternative to the kinetic description of plasmas, gases, and liquids. Their simulation requires fewer degrees of freedom than phase space models, yet it can still incorporate kinetic effects to a certain extent. To derive moment equations, we use a parameterization of the distribution function using centered moments, as proposed by Burby. This yields moment equations for which the parameterized distribution function exactly solves the hyperbolic conservation law. Similarly, a particle model is derived based on a parametrization of the distribution function using phase space moments. Finally, we present the application of the method to the non-relativistic and relativistic Vlasov--Maxwell equations.
