Structure Function Analysis of Long-Range Correlations in Plasma Turbulence

CX Yu,M Gilmore,WA Peebles,TL Rhodes
DOI: https://doi.org/10.1063/1.1583711
IF: 2.2
2003-01-01
Physics of Plasmas
Abstract:Long-range correlations (temporal and spatial) have been predicted in a number of different turbulence models, both analytical and numerical. These long-range correlations are thought to significantly affect cross-field turbulent transport in magnetically confined plasmas. The Hurst exponent, H—one of a number of methods to identify the existence of long-range correlations in experimental data—can be used to quantify self-similarity scalings and correlations in the mesoscale temporal range. The Hurst exponent can be calculated by several different algorithms, each of which has particular advantages and disadvantages. One method for calculating H is via structure functions (SFs). The SF method is a robust technique for determining H with several inherent advantages that has not yet been widely used in plasma turbulence research. In this article, the SF method and its advantages are discussed in detail, using both simulated and measured fluctuation data from the DIII-D tokamak [J. L. Luxon and L. G. Davis, Fusion Technol. 8, 441 (1985)]. In addition, it is shown that SFs used in conjunction with rescaled range analysis (another method for calculating H) can be used to mitigate the effects of coherent modes in some cases.
What problem does this paper attempt to address?