Low-rank approximation in the numerical modeling of the Farley-Buneman instability in ionospheric plasma
arXiv:1308.5952 · doi:10.1016/j.jcp.2014.01.029
Abstract
We consider the numerical modeling of the Farley-Buneman instability development in the earth's ionosphere plasma. The ion behavior is governed by the kinetic Landau equation in the four-dimensional phase space, and since the finite difference discretization on a tensor product grid is used, this equation becomes the most computationally challenging part of the scheme. To relax the complexity and memory consumption, an adaptive model reduction using the low-rank separation of variables, namely the Tensor Train format, is employed. The approach was verified via the prototype MATLAB implementation. Numerical experiments demonstrate the possibility of efficient separation of space and velocity variables, resulting in the solution storage reduction by a factor of order tens.
References in corpus (4)
- The density-matrix renormalization group in the age of matrix product states
- Alternating minimal energy methods for linear systems in higher dimensions. Part II: Faster algorithm and application to nonsymmetric systems
- Computation of extreme eigenvalues in higher dimensions using block tensor train format
- Quasioptimality of maximum-volume cross interpolation of tensors
Cited by in corpus (8)
- Parallel cross interpolation for high-precision calculation of high-dimensional integrals
- A semi-Lagrangian Vlasov solver in tensor train format
- A quantum-inspired method for solving the Vlasov-Poisson equations
- Tensor methods for the Boltzmann-BGK equation
- Parallel numerical tensor methods for high-dimensional PDEs
- A Parallel Low-Rank Solver for the Six-Dimensional Vlasov-Maxwell Equations
- Numerical solution of the Boltzmann equation with S-model collision integral using tensor decompositions
- Dynamical Tensor Train Approximation for Kinetic Equations