6.1.3. The UV Background

As in the case of the observed Infrared background, the observed diffuse UV background is dominated by dust scattering and interstellar emission in our own Galaxy. However, when the galaxy is subtracted out, a nearly-isotropic residual, with an amplitude of I ~ 100 photons s^-1 cm^-2 sr^-1 Å^-1 over the range 1300-2500 Å, is apparent (see Henry and Murthy 1993). Unfortunately, the calibration of this residual is very difficulty and the quoted amplitude is uncertain by a factor of 2-3. There are three plausible extragalactic sources for the observed residual: 1) diffuse thermal emission that would arise from material between the galaxies - the intergalactic medium (IGM) 2) the aggregate light from discrete sources such as UV bright QSOs and star forming galaxies; 3) radiation associated with the decay of some relatively long lived massive particle.

Although the primordial nucleosynthesis constrains on _b are fairly stringent, it is nevertheless important to observationally constrain the amount of material that might be contained in a hot intergalactic medium (IGM). The presence of a relatively warm (e.g., 10⁴ - 10⁶ K) dense IGM would manifest itself in the UV through the redshift blending of individual emission line features, principally Lyman from Hydrogen and He II 304. To date, the best UV probes of the nature of any "hot" IGM come from the myriad of QSO absorption line studies. This gives information on both the density of the IGM along various lines of sight out to redshift of z 4 and the state of its ionization. From these studies it is clearly known that the IGM contains some baryonic matter, mostly in the form of the ubiquitous "Lyman forest" clouds first characterized by Sargent et al. (1980). The source of ionization of these structures are most likely the QSOs themselves whom, possibly together with forming galaxies, produce the metagalactic UV flux.

In fact, the contribution of QSOs to the metagalactic UV flux can be directly estimated from a phenomenon known as the "proximity effect". This effect is the following: Imagine there is a string of Lyman forest clouds at various redshifts along the line of sight toward some distant QSOs. If one of these clouds is sufficiently close to the QSO itself, that cloud will see a source of ionizing flux that is above the metagalactic flux owing to its proximity to a single QSO. This extra UV flux from the QSO itself raises the overall state of ionization in the proximate clouds and hence reduces their ability to produce absorption lines. Since the UV flux from the QSO itself is directly observed, then the metagalactic UV flux can be inferred by comparing the UV absorption line strengths from those clouds that are near and far from the QSO. Over the redshift range 1.5-4, the level of this metagalactic UV flux is I _H ~ 10^-21 erg s^-1 cm^-2 sr^-1 Hz^-1.

This flux serves to ionize any intergalactic hydrogen or helium. These ionized gases cool via recombination line emission. The two dominant channels are the Lyman and He II 304 recombination lines. In this way, the photo-ionized gas acts as a photon energy down converter (e.g. continuum ionizing photons of wavelength less than 912 Å are down-converted into 1216 Å Lyman photons). Since the intensity of the recombination flux can never exceed that of the input ionizing flux, then individual line contributions to the total UV background can be estimated from the observed or inferred metagalactic UV flux over some redshift interval. In numerical terms, this can be expressed as :

(1)

where _h refers to Lyman limit photons. The (1 + z)⁴ term accounts for photon dilution effects as the radiation is redshifted to z = 0. Redshift-smeared Lyman recombination radiation from sources in the redshift range 0-0.6 would contribute to the 1200-2000 Å UV background. There are rather few QSOs in this redshift range and the available ionizing flux is down by 2 orders of magnitude relative to that quoted above for the redshift range 1.5-4. From equation 6.1, this yields an upper limit of 2 photons s^-1 cm^-2 sr^-1 Å^-1 which is only 2% of the observed extragalactic flux. The corresponding case of redshift-smeared He II 304 recombination radiation, which comes from the redshift range z = 3-5, suffers from more severe photon dilution. Using the nominal flux of I _H ~ 10^-21 erg s^-1 cm^-2 sr^-1 Hz^-1 in this redshift range yields less than 1 photons s^-1 cm^-2 sr^-1 Å^-1.

These simple calculations effectively demonstrate that recombination radiation from a dense IGM of hydrogen and helium does not effectively contribute to the observed extragalactic UV flux at z = 0. However, there is an alternate possibility that the IGM is shock heated and hence Hydrogen and Helium are collisionally ionized. This can be completely ruled out in the case of hydrogen as the total amount of neutral hydrogen in the IGM is strongly constrained by the lack of H I absorption troughs seen in the spectra of distant QSOs (historically this is referred to as the Gunn-Peterson test (see Giallongo et al. 1994; Fang and Crotts 1995). For the case of Helium, absorption troughs have now been detected toward at least two distant QSOs indicating the existence of some neutral intergalactic helium at redshifts z 3 (Davidsen et al. 1996). However, any 304 emission associated with the collisional ionization of this helium IGM would be rapidly damped out via the cumulative absorption by the "Lyman limit" hydrogen systems. We can thus confidently conclude that thermal emission from any IGM is a negligible component of the observed UV background. The most viable candidate for producing the background is therefore the integrated light from QSOs and star-forming galaxies as we dismiss the decaying particle hypothesis below.