Self-similarity, power-law scaling, and non-Gaussianity of turbulent fluctuation flux in a nonfusion magnetoplasma

Abstract

The statistical characteristics of turbulent particle flux Γturb due to the correlated fluctuations of density and E⃗×B⃗-drift velocity in the simple magnetized torus Thorello [S. Alba, M. Fontanesi, A. Galassi, C. Riccardi, and E. Sindoni, Phys. Essays 6, 225 (1993)] are studied. The focus is on the shape and the low-order moments (absolute width, rms width, skewness, and kurtosis) of the probability distribution functions (PDF), and on how they change as a function of time-resolution τ upon coarse-graining. The way the shape of the PDF and its moments change are closely linked to each other, and they lend support to the view of two distinct regimes, separated by τbreak≈50 μs. In both regimes, however, the statistical evidence for self-similarity, power-law scaling, and non-Gaussianity is rather mixed. The behavior at small τ is interpreted as due to the presence of smooth, positively correlated (and possibly coherent, i.e., long-lived) structures within the data, and the behavior at large τ as due to the standard (i.e., Gaussian, as opposed to the Levy–Gnedenko) Central Limit Theorem.