next up previous
Next: 3. The Magnetic Up: Tonguesbottles, and disconnected Previous: 1. Introduction

2. Opening Magnetic Fields

Coronal mass ejections (CMEs) [e.g., Gosling et al., 1974; Hundhausen, 1993, and references therein] are transient eruptions of coronal material and imbedded magnetic field loops which expand outward into interplanetary space. While none were occurring during the time of the image shown in Figure 1, CMEs are often detected in sequences of coronagraph images as bright loops rising up and expanding out of the coronagraph field of view (see, for example, Gosling et al. [1974]). Long lasting, bright radial structures called ``legs'' are often observed along the flanks of a CME; legs typically persist for many hours after the release of a CME. Since CMEs originate from regions of closed coronal magnetic fields that were not previously participating in solar wind expansion [ Gosling et al. 1976], legs have commonly been interpreted as evidence for continued magnetic connection of CMEs back to the Sun indicating that CMEs open new magnetic fields into interplanetary space. This interpretation of CME legs has recently been disputed by Kahler and Hundhausen [1992] who suggest that these bright structures may simply be coronal streamers containing magnetic neutral sheets and the higher density plasma associated with them.

Solar wind electron distributions are characterized by two components: a thermal population and a hotter ``halo'' population. Halo electrons usually have a beamed component directed along the local magnetic field; these beamed electrons do not interact much with the rest of the solar wind and can be treated essentially as test particles tracing out magnetic field topologies. Unidirectional halo electrons beamed outward from the Sun along the IMF indicate field lines that are tied back to a hot (10 K) corona at one end but not at the other (A in Figure 2).

In interplanetary space CMEs have been identified by a number of plasma and field signatures (summarized by Gosling [1990]). Of all of these signatures, halo electrons beaming both directions along the magnetic field (counterstreaming) seem to provide the most unambiguous indication of CMEs in the interplanetary medium [ Gosling, 1990; Gosling et al., 1992]. The reason is that just as a unidirectional beam directed outward along the IMF indicates a single connection back to a hot coronal source, beams directed both directions can best be interpreted as indicating connection to the Sun both ways along the IMF (B in Figure 2). Alternately, if magnetic reconnection pinches off a magnetic tongue to form a plasmoid, (C in Figure 2), it is possible that halo electrons already beaming both directions along the field could continue to circulate on such a structure for a long time, owing to their long interaction lengths. Phillips et al. [1992] find that asymmetries between the counterstreaming beams and variations in the field orientation across CMEs indicate that CMEs maintain at least some magnetic connection back to the corona. In addition, they documented the first examples of distributions which display two broad counterstreaming beams with an additional narrow beam superposed in the direction running most nearly outward from the Sun along the IMF. These distributions provide additional evidence for continued magnetic connection back to the Sun Phillips et al., 1992].

The other way that a subset of all CMEs is most commonly identified in interplanetary space is as ``magnetic clouds'' [ Klein and Burlaga, 1982; Zhang and Burlaga, 1988]. Magnetic clouds are defined by the combination of 1) large-scale smooth field rotations, 2) enhanced magnetic field magnitude, and 3) decreased plasma temperatures (e.g., the recent review by Burlaga [1991] and references therein). While only 10% of all CMEs identified with counterstreaming halo electrons fit the strict definition of magnetic clouds, roughly one third of all CMEs in interplanetary space exhibit at least the large, smooth field rotations characteristic of magnetic clouds [ Gosling 1990]. Such bundles of twisted, helical magnetic fields (also called flux ropes) probably arise from magnetic reconnection; Figure 3 schematically shows how reconnection between the lower parts of adjacent magnetic arches with sheared or offset foot points will form a flux rope-type topology and underlying magnetic loops [ Gosling 1993]. Of course this process would not be expected to occur only between a single pair of field lines (as shown for simplicity), but rather across a broad region, producing a long (and potentially magnetically complicated) flux rope-type CME (magnetic cloud), free to expand into interplanetary space, and an arcade of underlying coronal loops which remain tied to the Sun. Kahler and Reames [1991] used solar energetic particle observations inside magnetic clouds to show that these structures probably maintain magnetic connection to the Sun. These authors also suggested that clouds may occur on open field lines, tied to the Sun at only one end, however, Gosling et al. [1992] rebutted this latter suggestion.

Soft X-ray observations of the lower corona (as in Figure 1) indicate that roughly one third of all CMEs leave underlying arcades [ Sheeley et al., 1983]; this fraction is approximately the same as the fraction of all CMEs that have flux rope- or cloud-like properties in interplanetary space [ Gosling, 1993; Gosling et al., 1994]. One particularly clear case study was recently shown by Gosling et al. [1994] in which post CME loop formation observed with the Yohkoh soft X-ray imager [ Tsuneta et al., 1991] was correlated with the subsequent passage of a flux rope-type CME past Ulysses at 4.6 AU (astronomical units --- the distance from the Earth to the Sun). Motions of magnetic footprints down in the turbulent photosphere will naturally lead to shearing of the fields and can also bring the feet closer together increasing the likelihood of reconnection. In such a complicated configuration it is hard to imagine how all of the appropriate field lines could just happen to be positioned such that reconnection would form a fully detached plasmoid. Rather, it seems far more likely that some amount of magnetic connection is maintained for all CMEs and that some field lines are newly opened as CMEs expand outward through interplanetary space.



next up previous
Next: 3. The Magnetic Up: Tonguesbottles, and disconnected Previous: 1. Introduction



U.S. National Report to IUGG, 1991-1994
Rev. Geophys. Vol. 33 Suppl., © 1995 American Geophysical Union