Speaker
Description
The boundary of magnetically confined plasmas is in a strongly fluctuating state due to the propagation of large amplitude coherent structures. A stochastic single-point model based on the superposition of un-correlated pulses has been successfully validated across different reactors. We extend this model to describe two-point measurements. This model can describe experimental techniques such as gas puff imaging which are routinely performed in reactor shots. Within this framework, we develop methods to estimate model parameters, such as the pulse duration, asymmetry and distribution of delays. These methods can be used to obtain accurate estimations of the velocity of propagating structures. Methods traditionally used in the literature are based on maximizing the cross-correlation or the cross-conditional average. Using synthetic data, we compare our estimates with those obtained with traditional methods and show that our method is significantly less biased and can be used to obtain higher-order moments of the delays.
