No Arabic abstract
We present results from high-resolution hydrodynamic simulations of isolated SMC- and Milky Way-sized galaxies that include a model for feedback from galactic cosmic rays (CRs). We find that CRs are naturally able to drive winds with mass loading factors of up to ~10 in dwarf systems. The scaling of the mass loading factor with circular velocity between the two simulated systems is consistent with propto v_c^{1-2} required to reproduce the faint end of the galaxy luminosity function. In addition, simulations with CR feedback reproduce both the normalization and the slope of the observed trend of wind velocity with galaxy circular velocity. We find that winds in simulations with CR feedback exhibit qualitatively different properties compared to SN driven winds, where most of the acceleration happens violently in situ near star forming sites. In contrast, the CR-driven winds are accelerated gently by the large-scale pressure gradient established by CRs diffusing from the star-forming galaxy disk out into the halo. The CR-driven winds also exhibit much cooler temperatures and, in the SMC-sized system, warm (T~10^4 K) gas dominates the outflow. The prevalence of warm gas in such outflows may provide a clue as to the origin of ubiquitous warm gas in the gaseous halos of galaxies detected via absorption lines in quasar spectra.
The escape of cosmic rays from the Galaxy leads to a gradient in the cosmic ray pressure that acts as a force on the background plasma, in the direction opposite to the gravitational pull. If this force is large enough to win against gravity, a wind can be launched that removes gas from the Galaxy, thereby regulating several physical processes, including star formation. The dynamics of these cosmic ray driven winds is intrinsically non-linear in that the spectrum of cosmic rays determines the characteristics of the wind (velocity, pressure, magnetic field) and in turn the wind dynamics affects the cosmic ray spectrum. Moreover, the gradient of the cosmic ray distribution function causes excitation of Alfven waves, that in turn determine the scattering properties of cosmic rays, namely their diffusive transport. These effects all feed into each other so that what we see at the Earth is the result of these non-linear effects. Here we investigate the launch and evolution of such winds, and we determine the implications for the spectrum of cosmic rays by solving together the hydrodynamical equations for the wind and the transport equation for cosmic rays under the action of self-generated diffusion and advection with the wind and the self-excited Alfven waves.
We study the conditions for disk galaxies to produce superbubbles that can break out of the disk and produce a galactic wind. We argue that the threshold surface density of supernovae rate for seeding a wind depends on the ability of superbubble energetics to compensate for radiative cooling. We first adapt Kompaneets formalism for expanding bubbles in a stratified medium to the case of continuous energy injection and include the effects of radiative cooling in the shell. With the help of hydrodynamic simulations, we then study the evolution of superbubbles evolving in stratified disks with typical disk parameters. We identify two crucial energy injection rates that differ in their effects, the corresponding breakout ranging from being gentle to a vigorous one. (a) Superbubbles that break out of the disk with a Mach number of order 2-3 correspond to an energy injection rate of order 10^{-4} erg cm^{-2} s^{-1}, which is relevant for disk galaxies with synchrotron emitting gas in the extra-planar regions. (b) A larger energy injection threshold, of order 10^{-3} erg cm^{-2} s^{-1}, or equivalently, a star formation surface density of sim 0.1 solar mass yr^{-1} kpc^{-2}, corresponds to superbubbles with a Mach number sim 5-10. While the milder superbubbles can be produced by large OB associations, the latter kind requires super-starclusters. These derived conditions compare well with observations of disk galaxies with winds and the existence of multiphase halo gas. Furthermore, we find that contrary to the general belief that superbubbles fragment through Rayleigh-Taylor (RT) instability when they reach a vertical height of order the scale height, the superbubbles are first affected by thermal instability for typical disk parameters and that RT instability takes over when the shells reach a distance of approximately twice the scale height.
Accretion disk winds are thought to produce many of the characteristic features seen in the spectra of active galactic nuclei (AGN) and quasi-stellar objects (QSOs). These outflows also represent a natural form of feedback between the central supermassive black hole and its host galaxy. The mechanism for driving this mass loss remains unknown, although radiation pressure mediated by spectral lines is a leading candidate. Here, we calculate the ionization state of, and emergent spectra for, the hydrodynamic simulation of a line-driven disk wind previously presented by Proga & Kallman (2004). To achieve this, we carry out a comprehensive Monte Carlo simulation of the radiative transfer through, and energy exchange within, the predicted outflow. We find that the wind is much more ionized than originally estimated. This is in part because it is much more difficult to shield any wind regions effectively when the outflow itself is allowed to reprocess and redirect ionizing photons. As a result, the calculated spectrum that would be observed from this particular outflow solution would not contain the ultraviolet spectral lines that are observed in many AGN/QSOs. Furthermore, the wind is so highly ionized that line-driving would not actually be efficient. This does not necessarily mean that line-driven winds are not viable. However, our work does illustrate that in order to arrive at a self-consistent model of line-driven disk winds in AGN/QSO, it will be critical to include a more detailed treatment of radiative transfer and ionization in the next generation of hydrodynamic simulations.
Cosmic Rays escaping the Galaxy exert a force on the interstellar medium directed away from the Galactic disk. If this force is larger than the gravitational pull due to the mass embedded in the Galaxy, then galactic winds may be launched. Such outflows may have important implications for the history of star formation of the host galaxy, and in turn affect in a crucial way the transport of cosmic rays, both due to advection with the wind and to the excitation of waves by the same cosmic rays, through streaming instability. The possibility to launch cosmic ray induced winds and the properties of such winds depend on environmental conditions, such as the density and temperature of the plasma at the base of the wind and the gravitational potential, especially the one contributed by the dark matter halo. In this paper we make a critical assessment of the possibility to launch cosmic ray induced winds for a Milky-Way-like galaxy and how the properties of the wind depend upon the conditions at the base of the wind. Special attention is devoted to the implications of different conditions for wind launching on the spectrum of cosmic rays observed at different locations in the disc of the galaxy. We also comment on how cosmic ray induced winds compare with recent observations of Oxygen absorption lines in quasar spectra and emission lines from blank-sky, as measured by XMM-Newton/EPIC-MOS.
We study outflows driven by Active Galactic Nuclei (AGNs) using high- resolution simulations of idealized z=2 isolated disk galaxies. Episodic accretion events lead to outflows with velocities >1000 km/s and mass outflow rates up to the star formation rate (several tens of Msun/yr). Outflowing winds escape perpendicular to the disk with wide opening angles, and are typically asymmetric (i.e. unipolar) because dense gas above or below the AGN in the resolved disk inhibits outflow. Owing to rapid variability in the accretion rates, outflowing gas may be detectable even when the AGN is effectively off. The highest velocity outflows are sometimes, but not always, concentrated within 2-3 kpc of the galactic center during the peak accretion. With our purely thermal AGN feedback model -- standard in previous literature -- the outflowing material is mostly hot (10^6 K) and diffuse (nH<10^(-2) cm-3), but includes a cold component entrained in the hot wind. Despite the powerful bursts and high outflow rates, AGN feedback has little effect on the dense gas in the galaxy disk. Thus AGN-driven outflows in our simulations do not cause rapid quenching of star-formation, although they may remove significant amounts of gas of long (>Gyr) timescales.