This work continues earlier studies of the evolution of large-scale magnetic patterns which reached their climax in the formation of large complex active regions that produced numerous energetic white-light and proton-emitting flares (Bumba & Sýkora 1974, Bumba 1982, Bumba & Hejna, 1987). In each case the time-scale of the process was of the order of one year or more, and in all cases the local magnetic fields were transformed step by step from a relatively simple configuration into a more complex one. A number of processes were involved in this transformation: the appearance of new magnetic flux in active regions or in active-region complexes; the dissipation and changes in structure of older magnetic fields; the mutual interaction of all fields and their subsequent weakening as well as strenghtening; etc. All processes were found to be time antisymmetric; i. e., the growth phase was about 6 times longer than the declining phase. Similarly, the rapid final phase was accompanied by numerous flares and CMEs. Simultaneously, the entire area of complicated magnetic field patterns on the Sun, which at the phase of peak activity occupied as much as half of the solar surface, also disintegrated during this declining phase.
The persistence in the same longitude interval of the activity complex described in this paper suggests it was anchored to a rigidly rotating subsurface source of magnetic flux, which was not affected by the differential rotation, and which operated with maximum power through much of the complex's final phase. However, we have also noted significant changes in the local (possible local strenghtening of magnetic flux due to the strong motions, Bumba et al. 1995a, 1996) and large-scale magnetic topology that occurred during this period. At some point during this final period of maximum activity, the complex seems to become disconnected with this subsurface source, or the source itself has changed.
The evolution of this activity complex also parallels the evolution of large-scale magnetic fields studied during 1991 and 1992 (Bumba et al. 1995b). In both cases the disintegration of a large-scale magnetic complex led to the formation of a magnetically open region over the areas of formerly strong activity. In both cases, a large polar coronal hole penetrated into the area of the former activity complex and replaced it. From this we conclude that the development of a large coronal hole is a global process depending on the evolution of the global magnetic field. In addition, the evidence of this current study for the rigid rotation of magnetic boundaries of active-region complexes during the final stages of their evolution suggests "preparation" of the rigid boundaries of large coronal holes that penetrate low latitudes in the declining phase of the solar cycle. A well known example of this rigid rotation is the "boot of Italy" coronal hole observed by Skylab during 1973 (Krieger 1977). It should also be noted that the rigidly rotating pivot lines and pivot points discussed here appear to be associated with emerging flux regions and with the strengthening of this flux by other processes (Bumba et al. 1995a, 1996).
That substantial emerging flux may be needed during and immediately following the phase of peak activity and strong flaring, etc., is suggested by a recent study by Sudan & Spicer (1996). On the grounds of energetics and magnetodynamic stability, these authors argue that the energy for large flares must be emerging magnetic flux from subsurface regions, not magnetic energy stored in situ in the solar atmosphere. From this viewpoint, flares and other energetic manifestations of solar activity are mainly governed by the large-scale subsurface processes that determine magnetic flux generation and emergence, not by more local effects in the atmosphere itself. In further support of this picture, a possible driver for the "swelling" of helmet streamers often observed to precede CME eruptions could be the emergence of new magnetic flux from the subsurface regions (Hu, 1990), where similar energetic problems have been encountered (Low, 1993).
We also note there is evidence that the final opening into interplanetery space of the magnetic field lines in the region studied occurred well before its disconnection from a source of new flux, and even before the energetic event that announced the peak phase of activity, the flare and CME of July 9, 1982. This evidence is the difference between the weak radio emission recorded in June, 1989 from active region NOAA 3776 and the stronger, more typical radio emission from NOAA 3763 that was recorded earlier. Noting this, Bumba & Klvana (1997) report a corresponding change from a typically complex to a "twice bipolar" topological relationship between the magnetic fields of the activity complex and the global magnetic field, with the earlier case producing much more radio emission than the later, more dipole-like field.
At the end of the two-month period of peak activity in September 1982, the boundary of the large-scale region of positive magnetic field (the east edge of the leading portion of the global magnetic-bipolar structure) shifts eastward simultaneously with the simplification of the local magnetic field, "swallowing" the rest of the field into itself as it does so. This boundary is itself progressively transformed into, while merging with, the border of an expanding coronal hole, which started to grow earlier at higher latitudes, and continued to grow slowly to encompass lower latitude areas.
© European Southern Observatory (ESO) 1998
Online publication: December 16, 1997