Ellen G. Zweibel

On the origin of cosmic magnetic fields

We review the extensive and controversial literature concerning how the cosmic magnetic fields pervading nearly all galaxies and clusters of galaxies actually got started. Some observational evidence supports a hypothesis that the field is already moderately strong at the beginning of the life of a galaxy and its disc. One argument involves the chemical abundance of the light elements Be and B, while a second one is based on the detection of strong magnetic fields in very young high red shift galaxies.Since this problem of initial amplification of cosmic magnetic fields involves important plasma problems it is obvious that one must know the plasma in which the amplification occurs. Most of this review is devoted to this basic problem and for this it is necessary to devote ourselves to reviewing studies that take place in environments in which the plasma properties are most clearly understood. For this reason the authors have chosen to restrict themselves almost completely to studies of dynamos in our Galaxy. It is true that one can get a much better idea of the grand scope of galactic fields in extragalactic systems. However, most mature galaxies share the same dilemma as ours of overcoming important plasma problems. Since the authors are both trained in plasma physics we may be biased in pursuing this approach, but we feel it is justified by the above argument. In addition we feel we can produce a better review by staying close to that which we know best.In addition we have chosen not to consider the saturation problem of the galactic magnetic field since if the original dynamo amplification fails the saturation question does not arise.It is generally accepted that seed fields, whose strength is of order 10−20 G, easily spring up in the era preceding galaxy formation. Several mechanisms have been proposed to amplify these seed magnetic fields to a coherent structure with the microgauss strengths of the currently observed galactic magnetic fields.The standard and most popular mechanism is the α–Ω mean field dynamo theory developed by a number of people in the late sixties. This theory and its application to galactic magnetic fields is discussed in considerable detail in this review. We point out certain difficulties with this theory that make it seem unlikely that this is the whole story. The main difficulty with this as the only such amplification mechanism is rooted in the fact that, on galactic scales, flux is constant and is frozen in the interstellar medium. This implies that flux must be removed from the galactic discs, as is well recognized by the standard theory.For our Galaxy this turns out to be a major problem, since unless the flux and the interstellar mass are somehow separated, some interstellar mass must also be removed from the deep galactic gravitational well. This is very difficult. It is pointed out that unless the field has a substantial field strength, much larger than that of the seed fields, this separation can hardly happen. And of course, it must if the α–Ω dynamo is to start from the ultra weak seed field. (It is our philosophy, expressed in this review, that if an origin theory is unable to create the magnetic field in our Galaxy it is essentially incomplete.)Thus, it is more reasonable for the first and largest amplification to occur before the Galaxy forms, and the matter embedded in the field is gravitationally trapped. Two such mechanisms are discussed for such a pregalactic origin; (1) they are generated in the turbulence of the protogalaxy and (2) the fields come from giant radio jets. Several arguments against a primordial origin are also discussed, as are ways around them.Our conclusion as to the most likely origin of cosmic magnetic fields is that they are first produced at moderate field strengths by primordial mechanisms and then changed and their strength increased to their present value and structure by a galactic disc dynamo. The primordial mechanisms have not yet been seriously developed, and this preliminary amplification of the magnetic fields is still very open. If a convincing case can be made that these primordial mechanisms are necessary, more effort will of course be devoted to their study.

Magnetic Screening in Accreting Neutron Stars

We investigate whether the magnetic field of an accreting neutron star may be diamagnetically screened by the accreted matter. We assume the freshly accumulated material is unmagnetized and calculate the rate at which the intrinsic stellar magnetic flux is transported into it via Ohmic diffusion from below. For very high accretion rates (larger than the Eddington rate Edd), Brown & Bildsten have shown that the liquid ocean and outer crust of the neutron star are built up on a timescale much shorter than the Ohmic penetration time. We extend their work to lower accretion rates and calculate the resulting screening of the magnetic field. We find that the Ohmic diffusion and accretion timescales are equal for ≈ 0.1 Edd. We calculate the one-dimensional steady state magnetic field profiles and show that the magnetic field strength decreases as one moves up through the outer crust and ocean by n orders of magnitude, where n ≈ /0.02 Edd. We show that these profiles are unstable to buoyancy instabilities when B 1010-1011 G in the ocean, providing a new limit on the strength of the buried field. Our results have interesting implications for the weakly magnetic neutron stars in low-mass X-ray binaries. We find that magnetic screening is ineffective for < 10-2 Edd, so that, no matter how the accreted material joins onto the star, the underlying stellar field should always be evident. This is consistent with the fact that the only known persistently pulsing accreting X-ray millisecond pulsar, SAX J1808.4-3658, has an unusually low accretion rate of ~ 10-3 Edd. Although the simplified magnetic and accretion geometry we adopt here does not allow us to definitively say so, we are led to suggest that perhaps most of the weakly magnetic neutron stars in low-mass X-ray binaries have a screened magnetic field, explaining the lack of persistent pulsations from these sources. If screened, then the underlying field will emerge after accretion halts, on a timescale of only 100-1000 yr, set by the Ohmic diffusion time across the outer crust.

Magnetic Field Evolution in Neutron Star Crusts Due to the Hall Effect and Ohmic Decay

We present calculations of magnetic field evolution by the Hall effect and ohmic decay in the crust of neutron stars (NSs). In accreting NSs, ohmic decay is always the dominant effect because of the large resistivity. In isolated NSs with relatively pure crusts, the Hall effect dominates ohmic decay after a time tswitch 104 yr B, where B12 is the magnetic field strength in units of 1012 G. We compute the evolution of an initial field distribution by ohmic decay and give approximate analytic formulae for both the surface and interior fields as a function of time. Because of the strong dependence of tswitch on B12, early ohmic decay can alter the currents down to the base of the crust for B ~ 1011 G, neutron drip for B ~ 1012 G, and near the top of the crust for B 1013 G. We then discuss magnetic field evolution by the Hall effect. Several examples are given to illustrate how an initial field configuration evolves. Hall-wave eigenfunctions are computed, including the effect of the large density change across the crust. We estimate the response of the crust to the magnetic stresses induced by Hall waves and give a detailed discussion of the boundary conditions at the solid-liquid interface. Finally, we discuss the implications for the Hall cascade proposed by Goldreich & Reisenegger.

A MULTIWAVELENGTH STUDY OF M17: THE SPECTRAL ENERGY DISTRIBUTION AND PAH EMISSION MORPHOLOGY OF A MASSIVE STAR FORMATION REGION

We combine diffuse emission photometry from GLIMPSE and several other Galactic plane surveys covering near-IR through radio wavelengths to synthesize a global spectral energy distribution (SED) for the M17 complex. By balancing the integrated flux in the SED with the total bolometric luminosity of all known O and early B stars in the ionizing cluster, we estimate a distance to M17 of 1.6 kpc. At this distance, the observed total flux in the SED corresponds to a luminosity of 2.4 ± 0.3 × 106 L☉. We find that the SED from the H II region peaks at shorter wavelengths and has a qualitatively different shape than the SED from the photodissociation region (PDR). We find that polycyclic aromatic hydrocarbons (PAHs) are destroyed over a short distance or edge at the boundary of the H II region. We demonstrate that this PAH destruction edge can be located easily using GLIMPSE band-ratio images and confirm this using Spitzer IRS spectra. We investigate the relative roles of extreme ultraviolet (EUV) and X-ray photons in the destruction of PAHs, concluding that X-rays are not an important PAH destruction mechanism in M17 or, by extension, in any other Galactic H II region. Our results support the hypothesis that PAHs are destroyed by EUV photons within H II regions. PAHs dominate the mid-IR emission in the neutral PDR beyond the ionized gas.

Generation of the Primordial Magnetic Fields during Cosmological Reionization

We investigate the generation of magnetic fields by the Biermann battery in cosmological ionization fronts, using new simulations of the reionization of the universe by stars in protogalaxies. Two mechanisms are primarily responsible for magnetogenesis: (1) the breakout of ionization fronts from protogalaxies and (2) the propagation of ionization fronts through the high-density neutral filaments that are part of the cosmic web. The first mechanism is dominant prior to overlapping of ionized regions (z ≈ 7), whereas the second continues to operate even after that epoch. However, after overlap the field strength increase is largely due to the gas compression occurring as cosmic structures form. As a consequence, the magnetic field at z ≈ 5 closely traces the gas density, and it is highly ordered on megaparsec scales. The mean mass-weighted field strength is B0 ≈ 10-19 G in the simulation box. There is a relatively well-defined, nearly linear correlation between B0 and the baryonic mass of virialized objects, with B0 ≈ 10-18 G in the most massive objects (M ≈ 109 M☉) in our simulations. This is a lower limit, as lack of numerical resolution prevents us from following small-scale dynamical processes that could amplify the field in protogalaxies. Although the field strengths we compute are probably adequate as seed fields for a galactic dynamo, the field is too small to have had significant effects on galaxy formation, on thermal conduction, or on cosmic-ray transport in the intergalactic medium. It could, however, be observed in the intergalactic medium through innovative methods based on analysis of γ-ray burst photon arrival times.

Magnetic support of the optical emission line filaments in NGC 1275

The giant elliptical galaxy NGC 1275, at the centre of the Perseus cluster, is surrounded by a well-known giant nebulosity of emission-line filaments, which are plausibly in excess of 108 years old. The filaments are dragged out from the centre of the galaxy by radio-emitting ‘bubbles’ rising buoyantly in the hot intracluster gas, before later falling back. They act as markers of the feedback process by which energy is transferred from the central massive black hole to the surrounding gas. The mechanism by which the filaments are stabilized against tidal shear and dissipation into the surrounding extremely hot (4 × 107 K) gas has been unclear. Here we report observations that resolve thread-like structures in the filaments. Some threads extend over 6 kpc, yet are only 70 pc wide. We conclude that magnetic fields in the threads, in pressure balance with the surrounding gas, stabilize the filaments, so allowing a large mass of cold gas to accumulate and delay star formation.

The formation of sharp structures by ambipolar diffusion

The effect of ambipolar diffusion is investigated using simple numerical models. Examples are shown where sharp structures develop around magnetic nulls. In contrast to the case of ordinary diffusion, the magnetic field topology is conserved by ambipolar diffusion. This is demonstrated in an example where differential rotation winds up an initially uniform magnetic field and brings oppositely oriented field lines close together. It is argued that ambipolar diffusion produces structures of scales small enough for reconnection to occur.

A SURVEY OF EXTRAGALACTIC FARADAY ROTATION AT HIGH GALACTIC LATITUDE: THE VERTICAL MAGNETIC FIELD OF THE MILKY WAY TOWARD THE GALACTIC POLES

We present a study of the vertical magnetic field of the Milky Way toward the Galactic poles, determined from observations of Faraday rotation toward more than 1000 polarized extragalactic radio sources at Galactic latitudes |b| 77 ◦ , using the Westerbork Radio Synthesis Telescope and the Australia Telescope Compact Array. We find median rotation measures (RMs) of 0.0 ± 0. 5r ad m −2 and +6.3 ± 0. 7r ad m −2 toward the north and south Galactic poles, respectively, demonstrating that there is no coherent vertical magnetic field in the Milky Way at the Sun’s position. If this is a global property of the Milky Way’s magnetism, then the lack of symmetry across the disk rules out pure dipole or quadrupole geometries for the Galactic magnetic field. The angular fluctuations in RM seen in our data show no preferred scale within the range ≈0. ◦ 1t o≈25 ◦ . The observed standard deviation in RM of ∼ 9r ad m −2 then implies an upper limit of ∼1 μG on the strength of the random magnetic field in the warm ionized medium at high Galactic latitudes.

On the virial theorem for turbulent molecular clouds

An Eulerian, rather than Lagrangian, form of the virial theorem is derived for a turbulent, magnetized cloud embedded in a steady, turbulent, low-density intercloud medium. The role of turbulent pressure in cloud confinement is clarified, and it is shown that, in the absence of a magnetic field, a cloud can be at a somewhat lower pressure than the intercloud medium. Simple forms for the magnetic term in the virial equation are obtained. Radiation pressure is considered; its effects are relatively small under average conditions in the interstellar medium. Under typical conditions, external pressure and magnetic fields are shown to have a relatively small effect on virial estimates of the mass of self-gravitating clouds.

Magnetic reconnection in partially ionized gases

Magnetic field lines in a plasma reconnect at a rate scaled by the Alfven speed. In a partially ionized gas there are two natural Alfven speeds: one determined by the ionized mass density alone, which applies when ion-neutral friction is negligible, and one determined by the total mass density, which applies when ion-neutral friction is strong. When the ionization fraction is low, as in a dense molecular cloud, these two speeds differ by several orders of magnitude. Both time-dependent tearing modes and steady-state magnetic reconnection in partially ionized gas are considered, and the regimes in which the charged and neutral components are strongly, intermediately, and weakly coupled are delineated. Molecular clouds are probably in the intermediate regime, while reconnection in solar prominences probably has strong ion-neutral coupling. Reconnection proceeds more rapidly when coupling is not strong. 25 refs.