Gas flow into and out of galaxies
Standard cosmology predicts that at least 75% of the baryons are in the intergalactic medium rather than in galaxies. In rich clusters of galaxies the intergalactic medium is so hot and dense that it can be studied through its X-ray emission, but in the more typical environment of a galaxy such as the Milky Way, the gas is expected to be too rarefied to be observable with X-ray telescopes. It is vital to understand how this enormous pool of largely unobserved gas interacts with galaxies such as the Milky Way.
For over 40 years astronomers have been puzzled by the existence of 21-cm emission from neutral hydrogen (HI) along many lines of sight at unexpected velocities. The physical interpretation of this phenomenon is difficult because the 21-cm data give no indication of the distance to the emitting gas. Observations of absorption at ultraviolet wavelengths alow the distance to be probed if there is a suitable background source of ultraviolet radiation, such as a quasar or blue star. The number of such determinations has increased dramatically from 1999 when the FUSE satellite was launched. It turns out that many of the clouds are at distances of a few kpc above the Galactic disk.
In a parallel line of development, as 21-cm observatins of external galaxies weree extended to ever fainter flux levels, it was found that HI often extended kpc from the plane (e.g. Fraternali et al 2002). These two phenomena are surely the same but extensive modelling is required to demonstrate this and we are working on this (e.g. Fraternali & Binney 2005). There is some evidence that much of the explanar HI is shot up out of the plane by the explosive effects of star formation, but our modelling indicates that interaction with a warm diffuse medium that cannot be seen at 21-cm is also required.
Spitzer (1960) already inferred the value of the pressure that this medium exerts from measurements of interstellar absorption lines in the spectra of stars. The recent 21-cm data imply that the medium is slowly rotating. Measurements of heavy-element abundances in HI clouds suggest that the medium is substantially more metal-poor than the ejecta from the plane. The challenge is to synthesize all this information into a coherent overall picture.
Several lines of evidence indicate that "feedback" is an important aspect of galaxy formation. That is, when gas falls into a potential well and starts to form stars, black holes etc, the energy released in the formation of these objects blasts a lot of gas back out into intergalactic space. While it has long been anticipated that galaxy formation would involve some return of gas to intergalactic space, the efficiency of the return mechanism seems to be much higher than anticipated. We believe that the phenomenon of extraplanar HI is both an important clue in that its ejection can be studied observationally in great detail in nearby galaxies (including the Milky Way!) and a vital diagnostic of the Models of galaxy formation.