5.4. Indirect indicators of extragalactic magnetic fields
In addition to the yet-to-be-realized possibilities for the detection of intergalactic magnetic fields through low frequency synchrotron emission (section 4.3), some recent, and quite different types of observations might give powerful, even if indirect, evidence for the strength and extent of magnetic fields in the early Universe. In the following, four different lines of observational evidence are discussed which, when combined with some of the theoretical ideas reviewed elsewhere in this paper, have the potential for yielding new clues about the strength and presence of magnetic fields in the `quasar era' (redshifts ca 1 to 4) and possibly into the early galaxy formation era. It is at these epochs of the universe that we have a lamentable deficit of direct observational data.
5.4.1. The radio-far infrared correlation for galaxies in the nearby universe. Having been predicted by Harwit and Pacini (1975), a surprisingly robust correlation was discovered between the far infrared (FIR) emission from dust in starburst galaxies, and the co-extensive synchrotron emission which involves the interstellar magnetic field (Dickey and Salpeter 1984, Kronberg et al 1985, Helou et a1 1985, and others). Subsequent studies of this surprisingly tight, global correlation (Völk 1989, Helou and Bicay 1990, 1993) suggest that the interstellar gas and dust `knows' something about the magnetic field: empirically the correlation |B| n, where 1/3 2/3 over all galaxies (Helou and Bicay 1993) - n is the interstellar gas density associated with the far infrared emission.
A key observational fact for the purpose of this discussion is that, in starburst-like galaxies, the spectral density of the 40-100 µm FIR emission is much stronger, hence more easily visible to larger redshifts, than cm-wave synchrotron radiation. The coming generation of millimeter and submillimeter telescopes should be able to detect FIR emission at z 5 to 10, where the primeval analogues of starburst galaxies may be found, since this relatively strong emission would be redshifted to wavelengths at 1 mm. The apparently universal empirical relationship to magnetic fields would, given measured FIR flux densities, enable us to estimate the associated magnetic field strengths in primeval galaxies out to these large redshifts. Although indirect, this is one of the few ways by which we could infer magnetic field strengths associated with primeval galaxies during and before the `quasar epoch'.
5.4.2. Compton x-ray measurements on extragalactic sources of diffuse synchrotron radiation. Compton scattering of (especially mildly) relativistic electrons off microwave background photons to produce x-rays has been suggested and explored (Felten and Morrison 1963, Hoyle 1966, Rees 1967) in a variety of astrophysical contexts. Only recently has the quality of x-ray data `caught up' with the theory. The morphology and spectrum of 0.1-2 keV x-ray emission of the outflow gas around M82 (e.g. Schaaf et a1 1989) has been analyzed to estimate the inverse Compton scattering component of the synchrotron-emitting electron population off FIR photons generated in the nuclear starburst region. Similar comparisons have been made for x-ray-RM comparisons in galaxy clusters (cf section 4.l), and recent ROSAT x-ray images show that x-ray emission co-exists with radio emission in the outer lobes and hotspots of some extended radio galaxies - such as Cygnus A (Harris et al 1994).
Whereas Kim et al (1990) obtained a magnetic field estimate by comparing thermal x-ray emission with intracluster Faraday rotation, the x-ray data can also be analyzed in a different way; namely, to estimate that fraction of the x-ray emission which is due to inverse Compton scattering of microwave background photons off the same energetic electrons which produce the CO-spatial synchrotron emission. Since the photon density of the µwave background is well known, the emissivity of inverse Compton x-rays can be used to compute (or limit) the number density of relativistic electrons (ner0). Substituting this into the synchrotron emissivity equation (1.1) makes it possible to estimate |B|, or at least provide a lower limit to |B| corresponding to an upper limit on any inverse-Compton-generated x-rays. By isolating inverse Compton x-ray emission in the outer radio hotspots in Cygnus A, Harris et al (1994) have recently estimated the magnetic field strength in the Cygnus A hotspots to be 200 µG. Future generation x-ray images of extended radio lobes will permit this radio/x-ray method of magnetic field measurement to be made for other extragalactic radio sources - which can be seen over a substantial redshift range.
5.4.3. Inference of magnetic field strengths from highly redshifted metal-line QSO absorption lines. In section 3.1 we discussed recent evidence for the `seeding' of the IGM by vigorous outflow which has recently been discovered in starburst and other galaxies with active star-forming complexes. If, at the galaxy formation epoch during the first few percent of the Hubble time, starburst activity was ubiquitous in the compact universe in which galaxies were at only 10% of their current separation, a substantial fraction of the ICM at large redshifts might be filled by galactic wind plasma, consisting of magnetized gas which was enriched by heavy elements produced in the related supernovae.
This ejected, metal-enriched gas, if it exists far into galactic halos and beyond, could be the explanation of the surprising existence of heavier elements which have been seen in highly redshifted absorption lines in the spectra of some QSOs. If such enriched, magneto-ionic winds analogous to that of M82 (cf section 3.1) were produced in a compact `volcanic', early universe of starburst galaxies, then the dynamo action associated with these winds might have provided near-µG-level fields which seem to be ubiquitous at the present epoch wherever we find intergalactic gas. If this scenario is true, then observations of heavy elements in highly redshifted QSO absorption spectra of intervening gas clouds could possibly serve as effective tracers (even if not reliable strength indicators) of magnetic fields associated with the same systems.
5.4.4. Possible future use of cosmic rays to infer the geometry and maximum strength of intergalactic magnetic fields. The observed cosmic ray (CR) spectrum of light baryons, i.e. mainly protons, is found to extend to ca 1020 eV. Recent data from the Fly's Eye detector indicate that a chemical `switchover' occurs near 1018 - 1019 eV, in the sense that CRs above this range are mostly protons and helium, whereas heavy nuclei appear more predominant below ~ 1018 eV (Gaisser et al 1993). These most energetic CR protons are almost certainly extragalactic in origin. Biermann and Strittmatter (1987) and Rachen and Biermann (1993) argue that the outer hotspots of powerful FR II radio sources (cf section 3.2) are the most likely acceleration sites (by first order Fermi acceleration). If the universe is, as so far appears, filled with these energetic CR particles, then it is instructive to compare the average distance between the acceleration sites (~ 100 Mpc if luminous FR II radio sources are the origin of the CRs) with the gyroradius of 1020 eV protons in the presence of a widespread intergalactic magnetic field. A likely upper limit to the latter, discussed in section 5.1, is currently ~ 10-9 G. The gyroradius, rg, of a 1020 eV proton in a 10-9 G field is, interestingly, also of order 100 Mpc.
In the nearby universe, 1020 eV protons will travel in nearly straight lines for |BIG| < 10-9 G, which may be characteristic of the interiors (voids) of the large 100 Mpc bubbles which define the large scale distribution of optically visible matter. Around the periphery of the voids, i.e. the bubble `surfaces' where the galaxies lie, they will be substantially deflected if |BIG| = 10-7 -6 G, and possibly diffused in galaxy clusters and superclusters. Here, rg will be an order of magnitude or so smaller. In any case, rg, can be approximated to a mean free path CR(E, , BIG), which is given by
(Rachen and Biermann 1993) where is the CR particles's lifetime since acceleration to its initial (maximum) energy in units of 1017 s, E is in units of 1018 eV, and dbubble is the size of the cosmic bubble in Mpc. The assumptions we made about void interior and periphery field strengths, respectively, suggest that when the CR acceleration sites are identified (thus far none has been), a combination of position differences for different directions and distances could in future lead to evidence for intergalactic field strengths. Of particular interest will be the IG field strength within the voids, where we otherwise have little prospect of estimating or limiting the magnetic fields.
A relevant consideration is that high energy ( > 1010) CR protons will lose energy with cosmic time, principally due to -p interactions (which produce pions) with the cosmic microwave background radiation (MBR) (Greisen 1966, Zatsepin and Kuzmin 1966, Stecker 1968). Since the MBR photon density decreases with time in a cosmology-dependent way, the proton CR propagation (related to ) is complex (cf Hill and Schramm 1985, Berezinsky and Grigor'eva 1988). However the essence is that the attenuation length is very small, only 10 Mpc for > 1011, causing a cut-off in the energy spectrum of CR from cosmologically distant sources.
These dominant energy loss mechanisms cause the universe to be opaque to high energy proton CRs beyond a few tens of Mpc. They are of interest for this discussion for the following reasons: (i) they modify the CR energy spectrum in ways which can be observationally tested, (ii) they imply that direct sources of such CRs can only be identified in the local universe, and (iii) given this 'convenient' isolation of local extragalactic CR sources, the deflection and diffusion of high energy proton CRs in the local universe will depend on the intergalactic magnetic field, as discussed above. Future observations of the locations and isotropy of proton CRS will be influenced by the strength and morphology of the intergalactic field, so that discrepancies between the arrival directions of primary CR sources and their production sites might be established in future. Field strengths >> 10-9 G will cause significant deflection, whereas the converse is true if |BIG| << 10-9 G.