Rescaled range analysis of the asymmetry of solar activity

Previous studies of the north-south asymmetry of solar activity (e.g., Carbonell, Oliver, and Ballester, 1993; Oliver and Ballester, 1994) suggest that the asymmetry time series can be represented by means of a multicomponent model made up of a long-term trend, a weak sinusoidal component (with a period close to 12.1 years) and a dominant random process. Here, we have used the rescaled range analysis to study the valuation of the stochastic component of the asymmetry. To avoid the influence of the trend and the sinusoidal component on the result, we have removed both from the original time series. The value obtained for the Hurst exponent (0.717 ± 0.002) suggests that the non-periodic component is a correlated random process.