Annu. Rev. Astron. Astrophys. 1991. 29: 59-88 Copyright © 1991 by Annual Reviews. All rights reserved

6.3 Active Galactic Nuclei

The integrated light from active galactic nuclei (AGN), primarily QSOs, will contribute at some level to the diffuse far ultraviolet background. The intensity of this flux is of considerable interest because it is generally thought to be the factor that determines the ionization of intergalactic Lyman clouds and it may photoionize the Galactic halo. A number of authors have estimated the intensity of this flux (Jakobsen 1980; Sargent et al. 1980; Ikeuchi and Ostriker 1986; Bechtold et al. 1987; Martin and Bowyer 1989); the range of estimates is between 0.1 and 10 CU.

It is clear that the far ultraviolet flux produced by AGN will be too small to be directly disentangled from existing measurements. Future specialized searches may be capable of providing information on this topic.

6.4 Emission from an Intergalactic Medium

Early interest in the cosmic far ultraviolet background was motivated at least in part by its relevance to the existence of a cosmologically significant intergalactic medium (IGM). A high-temperature (> 10⁶ K) IGM would produce X-ray emission, and a lower-temperature (< 10⁴ K) medium would produce absorption effects in QSO spectra; observational constraints on these parameters were interpreted as ruling out these two scenarios, but the possibility of an intermediate-temperature IGM remained. An intermediate-temperature IGM would most likely be detected by its far ultraviolet emission; the most likely far ultraviolet radiation would be redshifted Lyman transitions of HI 1216 Å and HeII 304 Å from an ionized primordial gas (Field 1959; Kurt and Sunyaev 1967). In the early stages of galaxy formation this gas might be reionized collisionally by the passage of shocks or by interactions with energetic particles (Ginzberg and Ozernoy 1966; Weymann 1967), or by photoionization (Arons and McCray 1970). The emission spectrum depends upon the details and time scales of the reionization; in particular, collisional ionization of the gas produces much more far ultraviolet radiation than photoionization. Motivated by these suggestions, a variety of data was searched for indications of radiation from an intermediate-temperature IGM (Henry et al. 1978; Anderson et al. 1979; and Jakobsen 1980).

In the 1980s several groups developed more detailed models of far ultraviolet emission from a lukewarm IGM. Jakobsen (1980) showed that special conditions would be required if emission from this source were to contribute substantially (~ 300 CU) to the far ultraviolet background. In particular, collisional ionization and substantial clumping of the IGM would be required, and the thermal input must be carefully proscribed, or direct observational constraints would be violated.

Paresce, McKee, and Bowyer (1980) suggested that fast shocks and photoionization would be the most likely ionization mechanisms and these would produce prompt ionization with substantially less radiation. They examined the case in which the gas is sufficiently highly ionized that the dominant emission lines are produced primarily by recombination. In this scenario, the far ultraviolet emission from the IGM is ~ 10 CU.

Martin, Hurwitz, and Bowyer (1990) considered emission from an IGM for a variety of scenarios and compared their results with improved observational constraints provided by the X-ray background (Setti 1990), the lack of absorption troughs shortward of H Lyman continuum in the spectra of QSOs (Gunn and Peterson 1965), limits on emission in the far ultraviolet background as discussed in this article, and limits on spatial fluctuations in the far ultraviolet background (Martin and Bowyer 1989). They found that even their most optimistic scenarios, including collisional ionization and clumping of the IGM, violated observational constraints for an IGM with even a relatively low density (15% of the critical density of the universe). This result is to some extent model-dependent, but the models employed were sufficiently broad that either very special and unexpected conditions occurred in the early universe, or the far ultraviolet background has at most a component < 10 CUs from the IGM. A component this small will be difficult to identify.