Study On The Complexity Of The Long-Term Behavior Of The Solar Activity

Periodicity is a significant feature for long-term behavior of solar activity. But a variety of the periodicities has the different length and amplitude, such as the most representative of the solar cycle for 11 years. In addition, a variety of medium-term periodicities are intermittent, appears only in certain periods, and varying magnitudes. From the perspective of nonlinear dynamics, long-term solar activity(quasi-Week 11 years) showed a kinetic properties of low-dimensional strange attractor, is non-linear, chaotic. With development of nonlinear theory and nonlinear time series analysis methods being proposed, non-linear characteristics of the long-term solar activity(including quasi-11-year cycle, and the high frequency oscillations except 11-year cycle), such as the complexity research of it draws more and more attention.The multiscale sample entropy estimation method is applicable to full noise and short data. In this paper, sunspot numbers, solar filament number, solar coronal index and other data are chosen to study the different long-term solar activities and the complexity at different levels(the photosphere and the corona layer) of solar activities using multiscale sample entropy algorithm. The results of multiscale sample entropy show that: On average, the entropy value of sunspot number is higher than the coronal index showed higher complexity; the entropy value of solar filament is higher than sunspot number, showing the higher complexity, and the complexity of the high latitudes of the high latitude solar filament is the highest among them. The result of multiscale sample entropy calculation for each solar cycle of sunspots and coronal index showed that: entropy value of the 19th-23 rd solar cycle consistent with the results of the overall data, but the entropy value of the 18 th solar cycle shows abnormal behaviors that coronal index is higher than sunspots number, during the 18th-22 nd solar cycle, coronal index and sunspot number present the consistent entropy values, showing good correlation. Therefore, long-term solar corona layer is not necessarily more complicated than the photosphere, but with a different phenomenon related to solar activity.Combined with previous periodic results, the author believes that the complexity of the solar activities time series is related to how many types of the periodicities that the time series may contain. The more types of the periodicities the time sequence has, the higher regularity it reflects, relatively speaking, the lower the complexity of the data shows; on the contrary, the fewer types of the periodicities the time sequence has, the higher the complexity of the data shows. In addition, the multiscale sample entropy value of different solar activities showed that multiscale sample entropy method is applied to analyze long-term behavior of solar activity.