Generation of synthetic magnetized plasma turbulence by representation as a stochastic process

TL;DR

This paper introduces a statistical method to generate synthetic magnetized plasma turbulence profiles using a multivariate normal distribution based on the two-point covariance function, providing an efficient alternative to costly simulations.

Contribution

The authors propose a novel statistical approach to produce synthetic plasma turbulence data that accurately replicates real data's covariance structure, reducing computational costs.

Findings

Synthetic turbulence profiles match input data quantitatively.

Generated data are uncorrelated in time, suitable for static analysis.

Method offers a computationally efficient alternative to simulations.

Abstract

Plasma turbulence simulations are often computationally expensive with delicate numerical stability. Yet, long simulations are needed to generate uncorrelated turbulence data for studies such as microwave scattering through density perturbations. For this reason, alternative methods of producing accurate synthetic turbulence profiles via statistical methods is of interest. Such a method is proposed where the two-point covariance function of the desired turbulence is used to construct a multi-variate normal distribution. Sampling from this distribution produces random fields which are both qualitatively and quantitatively similar to the input turbulence data set. The resulting `synthetic' turbulent profiles are uncorrelated in `time' so it is useful only for scenarios that do not require such physical evolution.