Annu. Rev. Astron. Astrophys. 1996. 34: 155-206 Copyright © 1996 by Annual Reviews. All rights reserved

4. GALACTIC DYNAMO THEORY

We now discuss the mechanisms generating large-scale fields that have been presented in the previous section. We begin by considering first the small-scale magnetic fields.

4.1. Random Magnetic Fields

The interstellar medium is turbulent and thus any embedded magnetic field must have a random small-scale component. The presence of this component is crucial in all theories of large-scale dynamo action. There are several mechanisms that produce fluctuations in the interstellar magnetic fields: (a) tangling of the large-scale field by turbulence and from Parker and thermal instabilities, (b) compression of ambient magnetic fields by shock fronts associated with supernova remnants and stellar winds, and (c) self-generation of random magnetic fields by turbulence (small-scale dynamo). All of these mechanisms act together, and each imprints its own statistical properties onto the magnetic fields.

The available observational and theoretical knowledge of random magnetic fields and their maintenance in the ISM is rather poor. Instead, crude descriptions in terms of global quantities such as mean magnetic energy are usually applied. A widely used concept is that of equipartition between the magnetic and kinetic energy in the turbulence (Kraichnan 1965, Zweibel & McKee 1995), which implies that the rms random magnetic field strength is given by b B_eq (4 v²)^1/2, with v the rms turbulent velocity and the density. The equipartition value is significant in that the Lorentz force is expected to become comparable to the forces driving the turbulent flow as equipartition is approached. (This B_eq is not to be confused with the equipartition field strength in Sections 2 and 3, where equipartition refers to the estimated cosmic-ray energy density used to deduce the field strength from the synchrotron emission.) Interstellar turbulence is often treated as an ensemble of random Alfvén waves (McIvor 1977, Ruzmaikin & Shukurov 1982, McKee & Zweibel 1995) for which the equipartition holds exactly. Magnetic fluctuations are accompanied also by fluctuations in density (Armstrong et al 1995), so that other mechanisms, possibly nonpropagating fluctuations, must contribute to the interstellar turbulence (Higdon 1984). The random magnetic fields in the Milky Way are typically 4-6 µG (Ohno & Shibata 1993), close to B_eq.

Another component of the random magnetic field, one that is associated with interstellar (super) bubbles, is observed in the Milky Way (Heiles 1989, Heiles et al 1993, Vallée 1993). The magnetic field in HI shells, detected via the Zeeman effect, seems to be concentrated in filaments with the magnetic pressure larger than the gas pressure. The field strength in magnetic bubbles around OB associations, as obtained from Faraday rotation measurements, follows the density dependence b expected for a shocked medium.

The small-scale dynamo (Kazantsev 1968, Meneguzzi et al 1981) must be an important source of interstellar random magnetic fields (Sokoloff et al 1990). A distinctive feature of this component of the interstellar field, item (b) above, is that it is organized in intermittent magnetic ropes of small filling factor and lengths comparable to the correlation length of the turbulence (50-100 pc). The rms strength of the magnetic fluctuations generated by this mechanism is possibly close to the equipartition value, but the field within the filaments may be significantly higher (Belyanin et al 1993). For example, three-dimensional simulations of convective small-scale dynamo action at magnetic Reynolds numbers of about 1000 (Nordlund et al 1992) give B_rms = 0.4B_eq and B_max = 3B_eq. We note that in the interstellar gas of elliptical galaxies a small-scale dynamo may be the only source of magnetic fields, generating random fields of µG strength and a few hundred parsecs in scale (Moss & Shukurov 1996).