Astron. Astrophys. 322, 311-319 (1997)

1. Introduction

Solar activity study has been developed in several ways over the years. Among them, the sunspot variability has been greatly investigated because data go back to the year 1610. The long-term spot series shows an average cycle length of 11.1 years (Schwabe cycle), with a standard deviation of 0.26 year. However, Maunder's Minimum (interval with quasi-suppressed spot appearance: 1645-1705; Ribes & Nesme-Ribes 1993; Hoyt, Schatten & Nesmes-Ribes 1993; Eddy 1976) and several ancient periods inferred from solar proxy-data suggest, from the reduced activity, that the Sun exhibits quasi-periodic or intermittent behaviour. The former interpretation allows us to ascribe activity cycle anomalies to a stochastic process, while the latter pertains to the deterministic chaos. The matter is still controversial (e.g., Weiss 1990; Price et al. 1992; Rozelot 1995, among others). Hence, investigation of solar activity phenomena in terms of stochastic behaviour cannot be rejected and offers useful clues for the understanding of solar-type stars.

Several aspects of the 11-year solar-activity cycle were examined in the past. Without reviewing this long series of works, we only notice that, by examining measurements of the green line (Fe XIV, 5303 ) intensity of the solar corona ordered in heliographic latitudinal belts, Gnevyshev (1963) claimed the existence of a dual-peak structure for the coronal activity maximum of the cycle. The first peak was observed near the sunspot number maximum (1957) and involved the medium and high heliographic latitudes (maximum intensity at ); the second one occurred about two years later at low latitudes only (maximum at ). Similar results from other activity parameters (such as radio emission, sunspot areas, chromospheric and proton flares) and extended to past cycles reinforced the idea that, during each 11-year cycle, there is a splitting of the activity maximum in all the solar atmospheric layers. Two waves of activity, partly superimposed in time, were invoked (Gnevyshev 1967, 1977 and references therein) to explain this long-term trend.

In a review paper by Sýkora (1980) the isophote charts of the green corona brightness (derived from homogenized data of various coronal observatories) were published for cycles 18 to 20. It was easier to describe the outstanding coronal activity in terms of large well-isolated impulses (not necessarily in phase in both solar hemispheres) rather than in terms of two activity waves. The phenomenon was connected with the processes involved in the development of large and complex active regions (Bumba & Howard 1965). Moreover, Antalová & Gnevyshev (1983) published similar charts for sunspot areas (1874-1976 years), recognizing the possible existence of more peaks during a cycle (often a third peak). On the other hand, Mikhailutsa & Gnevyshev (1988) showed that for cycle 21 there exists a link between the double peak of the green corona activity cycle and the coronal magnetic energy maxima. The same cycle was investigated by Obridko & Shelting (1992), showing a relationship between the cyclic variation of the global heliomagnetic field (see also Sect. 5) and the one of several solar-geophysical parameters.

Moreover, recent studies on cosmic rays (Nagashima et al. 1991; Storini 1995; Storini & Pase 1995; Sýkora & Storini 1996; Storini et al. 1996) and geomagnetic activity (Gonzalez et al. 1990; Clúa de Gonzalez et al. 1994) recall the reliability of Gnevyshev's dual-peak in the heliospheric parameters.

Hence, the necessity for a better understanding of the 11-year cycles is emerging not only for solar and stellar structure studies but also for solar-terrestrial physics. This paper describes part of an ongoing investigation into the subject looking at the Sun as a star, i.e. neglecting the latitudinal distribution of activity centers (see Feminella & Storini 1996 for preliminary results).

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998
helpdesk.link@springer.de