Bayesian learning of impurity transport in tokamak high confinement regimes

Francesco Sciortino


Thursday, March 22, 2018



PSFC Student Seminars

The dynamics of impurities in high confinement scenarios rule out operational regimes where excessive power radiation or fuel dilution occurs. Understanding the mechanisms underlying impurity transport is therefore imperative to predict the behavior of future devices. Such effort also provides a compelling pathway for gyrokinetic model validation, where accurate estimation of uncertainties is paramount. Recent advances in Bayesian inference methods and the availability of extensive computational resources suggest new approaches to the evaluation of radial profiles of impurity transport coefficients from experimental data. This talk will introduce experimental scenarios and Bayesian techniques of interest, reporting results on EDA H-mode Alcator C-Mod plasmas.