Collaborative Research: Wave-Particle Interactions in 3D Magnetic Fields

About This Grant

Wave-particle interactions are a fundamental process underlying phenomena across the plasma universe, from laboratory plasmas to the magnetosphere. Understanding how energetic particles interact with waves in space and laboratory plasmas has the potential to improve our ability to protect satellites, design cleaner energy sources, and develop technologies that rely on controlling high-temperature plasmas. This award supports a collaboration between Columbia University, West Virginia University, and New York University to study how modulations of the background magnetic fields can impact the interactions between energetic particles and plasma waves. Machine learning techniques will be leveraged to discover simplified models that capture the relevant dynamics. In addition to advancing science, this project will support the training of students and early-career researchers, develop interactive classroom tools for K-12 and graduate education, and promote open, accessible science through videos, software, and tutorials. This project will bring together expertise from energetic particle dynamics in magnetic confinement fusion, radiation belt electron transport, and data-driven reduced models to address two fundamental questions: How are resonant wave-particle interactions (WPI) modified by three-dimensional (3D) structure of magnetic fields? and How do 3D magnetic fields modify wave-induced particle transport? These questions will be addressed using two model problems: resonant interaction of energetic particles with Alfvén waves and transport of radiation belt electrons by ultra low frequency (ULF) waves. The project will develop a reduced particle-based simulation framework to address these questions, taking advantage of the separation of timescales between the background evolution and resonant population evolution. This analysis will be complemented by data-based development of reduced-order models of WPI. An interpretable machine learning paradigm, sparse identification of nonlinear dynamics (SINDy), will be used to discover reduced models for particle transport due to WPI and 3D fields. These reduced transport models will fill the gap between quasilinear diffusion coefficients and particle tracing simulations, while also informing global magnetospheric modeling, where a neural network with an autoencoder architecture will be used to identify a nonlinear low-dimensional latent space where the nonlinear behavior of WPI can be mapped. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.