Development of Method for Determining Impurity Transport Coefficients in Tokamak Plasmas

To analyze impurity transport, it is necessary to determine the profiles of transport coefficients for impurities. In this study, an algorithm has been developed to determine transport coefficients and the accuracy has been validated using the phantom transport coefficients. The developed algorithm uses reconstruction methods to minimize differences between the simulated and measured impurity density profiles for continuous time series. Philips-Tikhonov (PT) regularization and Gaussian process (GP) regression are used to determine the transport coefficient in this work. To validate the accuracy of the developed algorithm, phantom transport coefficients of Ar and Ne obtained from the JET experiments have been utilized. Both PT-based and GP-based methods achieved root-mean-square errors of less than 1% for both the Ar and Ne cases. Moreover, the calculation time of the algorithm was about a few seconds, enabling fast analysis of impurity transport. The developed algorithm has been implemented into the KAIST impurity modelling code to investigate impurity transport in KSTAR plasmas, together with an infrared imaging video bolometer (IRVB) diagnostic for monitoring impurity density evolution. Using the determined transport coefficients from the krypton seeding experiment, the simulation yielded a comparable result to the measured density evolution. This outcome provides further evidence for the effectiveness of the developed algorithm in analyzing impurity transport in tokamak devices.