Data-driven discovery of reduced plasma physics models from fully-kinetic simulations

At the core of some of the most important problems in plasma physics — from controlled nuclear fusion to the acceleration of cosmic rays — is the challenge to describe nonlinear, multi-scale plasma dynamics. The development of reduced plasma models that balance between accuracy and complexity is critical to advancing theoretical comprehension and enabling holistic computational descriptions of these problems. In this talk, I will discuss how techniques from statistical and machine learning are offering new ways of inferring reduced plasma physics models from the increasingly abundant data of plasma dynamics produced by experiments, observations and simulations. In particular, I will focus on how sparse regression techniques can be used to infer interpretable plasma physics models (in the form of nonlinear partial differential equations) directly from the data of fully-kinetic particle-in-cell (PIC) simulations. I will demonstrate the potential of this approach by recovering the fundamental hierarchy of plasma physics models — from the kinetic Vlasov equation to single-fluid magnetohydrodynamics — based solely on data of complex plasma dynamics from first-principles PIC simulations. I will give some perspectives about how this data-driven methodology offers a promising new tool to accelerate the development of reduced theoretical models of complex nonlinear plasma phenomena and to design computationally efficient algorithms for multi-scale plasma simulations.