Bayesian Inference of H-mode Impurity Transport in Alcator C-Mod

The investigation of impurity transport offers a compelling pathway for the validation of turbulence models, in particular through the computation of radial profiles of impurity transport coefficients. Such validation effort requires a rigorous computation of uncertainties, which we approach in a Bayesian framework using Gaussian Process Regression and the MultiNest algorithm to sample from multi-dimensional, potentially multi-modal posterior distributions. For the first time, we applied such analysis to determine experimental impurity transport in an Alcator C-Mod EDA H-mode plasma, where impurity confinement times are expected to be larger than in previously analyzed L-mode conditions. Using spatially resolved measurements of Ca +18 provided by an X-ray Imaging Crystal Spectrometer (XICS) and the STRAHL impurity transport code, radial profiles of experimental diffusion and convection coefficients were obtained following the injection of impurities via Laser Blow-Off (LBO). These results constitute the first steps towards constructing an impurity transport experimental database that will be used to provide constraints for gyrokinetic model validation.