ECCOMAS 2024

Zooming into Kinetic Equations using the Characteristic Mapping Method

  • Krah, Philipp (Aix-Marseille Universitè)
  • Yin, Xi-Yuan (Universitè de Lyon)
  • Bergmann, Julius (Aix-Marseille Université)
  • Nave, Jean-Christophe (McGill University)
  • Schneider, Kai (Aix-Marseille Université)

Please login to view abstract download link

Our work introduces an efficient semi-Lagrangian characteristic mapping method for solving kinetic equations with high precision on coarse grids. Its ability is demonstrated for the one+one-dimensional Vlasov–Poisson equation. This method employs a numerically evolved flow map utilizing a gradient-augmented level-set technique, achieving exponential resolution within linear time. We demonstrate the method’s global third-order convergence in space and time while rigorously evaluating its conservation properties. Benchmarking against established test cases such as Landau damping and the two-stream instability showcases the effectiveness of our approach, highlighting its superiority over classical pseudo-spectral methods. Moreover, we illustrate the method’s exceptional capability to transcend the limits of currently available schemes by showcasing its extreme fine-scale resolution features. Our method stands as a promising advancement in addressing the computational challenges associated with kinetic theories, offering a pathway towards unprecedented precision in simulations.