Articles | Volume 8, issue 4/5
https://doi.org/10.5194/npg-8-313-2001
https://doi.org/10.5194/npg-8-313-2001
31 Oct 2001
31 Oct 2001

A study of magnetic fluctuations and their anomalous scaling in the solar wind: the Ulysses fast-latitude scan

c. Pagel and A. Balogh

Abstract. The solar wind is a highly turbulent and intermittent medium at frequencies between 10-4 and 10-1 Hz. Power spectra are used to look at fluctuations in the components of the magnetic field at high frequencies over a wide range of latitudes. Results show steady turbulence in the polar regions of the Sun and a more varied environment in the equatorial region. The magnetic field fluctuations exhibit anomalous scaling at high frequencies. Various models have been proposed in an attempt to better understand the scaling nature of such fluctuations in neutral fluid turbulence. We have used the Ulysses fast latitude scan data to perform a wide ranging comparison of three such models on the solar wind magnetic field data: the well-known P model, in both its Kolmogorov and Kraichnan forms, the lognormal cascade model and a model adapted from atmospheric physics, the G infinity model. They were tested by using fits to graphs of the structure function exponents g(q), by making a comparison with a non-linear measure of the deviation of g(q) from the non-intermittent straight line, and by using extended self similarity technique, over a large range of helio-latitudes. Tests of all three models indicated a high level of intermittency in the fast solar wind, and showed a varied structure in the slow wind, with regions of apparently little intermittency next to regions of high intermittency, implying that the slow wind has no uniform origin. All but one of the models performed well, with the lognormal and Kolmogorov P model performing the best over all the tests, indicating that inhomogeneous energy transfer in the cascade is a good description. The Kraichnan model performed relatively poorly, and the overall results show that the Kraichnan model of turbulence is not well supported over the frequency and distance ranges of our data set. The G infinity model fitted the results surprisingly well and showed that there may very well be important universal geometrical aspects of intermittency over many physical systems.