Microtearding mode study in NSTX using machine learning enhanced reduced model

• DOI: 10.48550/arxiv.2304.08982
• Publication Date: 2023-01-01
• Authors: Curie, Max T.; Larakers, Joel; Parisi, Jason; Staebler, Gary; Munaretto, Stefano; Guttenfelder, Walter; Belli, Emily; Hatch, David R.; Lampert, Mate; Avdeeva, Galina; Neiser, Tom; Smith, Sterling; Diallo, Ahmed; Nelson, Oak; Kaye, Stanley; Fredrickson, Eric; Manela, Joshua M; Lei, Shelly; Halfmoon, Michael; Tennery, Matthew M; Hassan, Ehab
• Keywords: Plasma Physics (physics.plasm-ph); FOS: Physical sciences; Physics - Plasma Physics

This article presents a survey of NSTX cases to study the microtearing mode (MTM) stabilities using the newly developed global reduced model for Slab-Like Microtearing modes (SLiM). A trained neutral network version of SLiM enables rapid assessment (0.05s/mode) of MTM with $98\%$ accuracy providing an opportunity for systemic equilibrium reconstructions based on the matching of experimentally observed frequency bands and SLiM prediction across a wide range of parameters. Such a method finds some success in the NSTX discharges, the frequency observed in the experiment matches with what SLiM predicted. Based on the experience with SLiM analysis, a workflow to estimate the potential MTM frequency for a quick assessment based on experimental observation has been established.