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

6.4. Contrast Structures

Suppose that the seed magnetic field in one part G₁ of a thin galactic disk has approximately the form of a growing eigensolution, while in another part G₂ the seed magnetic configuration is close to the same eigensolution, but with the opposite sign. After some time, advection and diffusion will bring these regions of oppositely directed magnetic fields into contact. The neutral surface at the boundary of these regions will move due to diffusion and advection, so the final stage of magnetic field evolution will be determined by magnetic field propagated, say, from the part G₁. The motion of the neutral magnetic surface is governed by the competition between advection and diffusion of field from G₁ towards G₂ and vice versa. Provided the nonlinear stage of magnetic field evolution begins before the field attains the form of the leading eigensolution, these two can balance each other. This balance is possible only if the neutral surface is at some special location in the galactic disk; then a long-lived magnetic structure appears (Belyanin et al 1994). This type of nonlinear solution of the dynamo equations is known as a contrast structure. The thickness of the transition region between G₁ and G₂ is approximately the disk thickness, and its lifetime can even be as long as the diffusion time along the disk, R² / _t ~ 10¹¹ yr. Inside the contrast structure, annihilation of the oppositely directed magnetic fields is balanced by generation and advection, similar to a solitonís behavior in the nonlinear wave equation.

Contrast structures in purely axisymmetric disks are expected to be most often axisymmetric, because they are not affected by differential rotation. In the Milky Way, such axisymmetric contrasting structures can survive until today, and they may be identified with the reversals discussed in Section 3.8.2 (Poezd et al 1993). Contrast structures supported by nonaxisymmetric velocity and density distributions might explain the dominance of BSS in some galaxies (Moss et al 1993b; D Moss, in preparation; A Bikov et al, in preparation).