Dynamic Mode Decomposition for Prediction of Kinetic Plasma Behavior

We present a reduced-order model based on the dynamic mode decomposition (DMD) for capturing the essential dynamics of kinetic plasma simulation. We apply the proposed methodology to the case of an accelerated electron beam propagation inside a cavity in the presence of a transverse magnetic field. The full order model utilizes a local and explicit finite-element time-domain (FETD) based electromagnetic particle-in-cell (EM-PIC) algorithm for simulating the beam. We employ DMD to extract dominant features that help us predict the future distribution of self electric field. The reduction achieved by this analysis can be instrumental in reducing the computation cost of complex plasma simulations.