alternate text

Simulated galactic outflow, transporting metals into the intergalactic medium. The galaxy sits at the convergence of intergalactic filaments (light blue). The composite image shows the gas density (in blue) and the metallicity distribution (in orange). Note how the metal rich gas avoids the filaments and escapes from the higher density regions into the voids, choosing the path of least resistance. This is one of the reasons why it is difficult for galactic winds to disturb the general filamentary intergalactic medium. Observations of QSO metal absorption systems indeed appear to show more widespread metal enrichment than such late time winds can produce. In reality, the blue filaments are themselves already enriched, suggesting that these metals originate in an earlier phase of galaxy formation and, at the time we observe them, are falling in rather than coming out of those galaxies. In addition, it is not clear whether such long range winds as assumed in the above model really exist. (simulation and figure by D. Kawata).


















Galactic Outflows

The intergalactic medium shows signs of having been enriched by metals even at the highest redshifts currently observable (z~6; e.g. Becker et al 2009). Where do these metals come from ? Have galaxies been stripped of their metal-rich gas by collisions or tidal interactions with each other, or by ram-pressure when moving through the ambient medium ? Or have they lost their gas through outflows, driven by active galactic nuclei, hot stellar winds, or supernova explosions ?

Spectra of star-forming galaxies typically show outflows of gas with velocities of sometimes hundreds of km/s. As these outflows are invariably measured in absorption in the spectra of the stars in those galaxies, it is not possible to directly infer their spatial range, i.e., whether the gas really gets out of the galaxy or remains confined within its interstellar medium. In the early 2000s the idea that the gas may be capable of leaving the galaxy gained popularity and papers appeared that ascribed the origin of the metal-enriched gas seen within hundreds of kiloparsecs of bright star-forming ("Lyman break") galaxies to outflows driven by those same starbursts. A survey of high ionization (OVI) QSO absorption lines with Keck HIRES by Simcoe, Sargent, and Rauch (2002) suggested that Lyman break galaxies may be surrounded by highly ionized gas out to a radius of 50 kpc, at least consistent with gas possibly expelled by galactic outflows. Indirect evidence for the IGM being repeatedly "stirred" by some process comes from observations of the kinematics of metal enriched gas clouds (Rauch et al 2001; see also the research link "IGM on small scales"). In particular the properties of CIV absorption systems, likely to arise in the outskirts of gaseous galactic halos, suggest that the turbulent energy of those systems is being replenished repeatedly. This may be consistent with the action of winds, but the turbulence could also be caused during gravitational infall of gas, by tidal interaction or during a merger with another galaxy, or simply whenever a galaxy plows through the intergalactic medium. Strong correlations between the column density of gas observed in lines-of-sight near star-bursting galaxies, and the distance from those galaxies have led to suggestions that there is a "circumgalactic medium", shaped by gas blown out of those same galaxies to hundreds of kpc.

However, the idea that the metal rich gas within hundreds of kpc of star-forming galaxies has been ejected from the very same galaxies has remained controversial. To start with, it requires the existence of high redshift galactic winds with a spatial range much further than that of actual galactic winds as observed at low redshift. In addition, the quiescence of the general intergalactic medium as seen when comparing close lines of sight to background QSOs limits the volume that could have been affected by recent outflows (Rauch et al 2001). Moreover, many of the properties of the absorption systems seen in high redshift QSOs are consistent with arising in gas that has been polluted by early galaxies pre-dating the star-forming objects in whose vicinity the gas is observed later (e.g., Rauch, Haehnelt and Steinmetz 1997). Thus, the association between massive galaxies and metal-enriched gas could be mostly representing correlation rather than causation., i.e., gas and galaxies are ending up in the same dark matter gravitational potential wells. Recent studies (Turner et al 2017) of metal absorption near Lyman break galaxies (prime candidates for galaxies driving galactic outflows into the IGM), suggest that the motions of the enriched gas surrounding those galaxies are dominated by gas flowing in, not out of them.

Daisuke Kawata and I have tried to reproduce metal absorption systems near star-forming galaxies with models of (hypothetical) strong local galactic winds. We found that the observational properties of CIV absorption systems (one of the most ubiquitous species) cannot be matched with late winds alone and requires an earlier widespread phase of metal enrichment. The actual absorption line signature of winds turns out to be rather subtle, and even extremely strong winds would not be able to make a big dent in the appearance of the general intergalactic medium. The winds were found to mostly take the path of least resistance and escape perpendicular to intergalactic medium filaments into the voids, where the metal enriched outflow would be very hard to see by any means.

The question as to how the metals get from the early galaxies to the IGM remains unclear. More moderate outflows from low mass galaxies at very high redshift and the subsequent Hubble expansion of the enriched gas are likely to be important. The difficulty of simulating realistic winds in a cosmological context, and the poorly investigated effect of other mass loss phenomena like ram pressure and tidal stripping on the level of individual galaxy mergers do not yet allow us to fully understand the relative importance of these processes and their cosmic timing.

Publications:
Becker, G. D., Rauch, Michael, Sargent, W. L. W., 2009, ApJ, 698, 1010

Rauch, M., Sargent, W.L.W., Barlow, T.A.: Small-Scale Structure at High Redshift. II. Physical Properties of the C IV Absorbing Clouds, Astrophysical Journal, 554:823-840, 2001

Simcoe, R.A., Sargent, W.L.W., Rauch, M.: Characterizing the Warm-Hot Intergalactic Medium at High Redshift: A High-Resolution Survey for O VI at z = 2.5 , Astrophysical Journal, 578:737-762, 2002

Kawata, D., Rauch, M.: Galactic Wind Signatures around High-Redshift Galaxies, Astrophysical Journal, 663:38, 2007