Introduction

There are a number of studies claiming that dust in foreground galaxies has a substantial effect on the colours and counts of optically-selected quasars (Ostriker & Heisler 1984; Heisler & Ostriker 1988; Fall & Pei 1992 and Wright 1990). It is estimated that at least 50% of bright quasars at redshifts may be obscured by foreground galactic dust and hence missing from optical samples. These studies assumed that dust was confined only within the visible extent of normal massive galaxies. However, distant populations such as faint field galaxies and quasars may also be observed through foreground diffuse dust distributions. Such distributions may be associated with galaxy clusters and extended galactic haloes.

A truly diffuse, intergalactic dust component is ruled out based on the counts of quasars and reddening as a function of redshift (eg. Rudnicki 1986; Ostriker & Heisler 1984). Such observations indicate that if a significant amount of dust exists, it must be patchy and diffuse with relatively low optical depth so that quasars will appear reddened without being removed from flux-limited samples.

Galaxy clusters provide a likely location for `large-scale' diffusely distributed dust. Indirect evidence is provided by several studies reporting large deficits of distant quasars or clusters of galaxies behind nearby clusters (Boyle et al. 1988; Romani & Maoz 1992 and references therein). These studies proposed that extinction by intracluster dust was the major cause. Additional evidence for diffuse dust distributions is provided by observations of massive local galaxies where in a few cases, dust haloes extending to scales kpc have been confirmed (Zaritsky 1994 and Peletier et al. 1995).

Does uniformly distributed dust really exist in the intergalactic medium (IGM)? Galactic winds associated with prodigious star formation at early epochs may have provided a likely source of metal enrichment and hence dust for the IGM (eg. Nath & Trentham 1997). Observations of metal lines in Ly- absorption systems of low column density () indeed suggest that the IGM was enriched to about by redshift (Womble et al. 1996; Songaila & Cowie 1996). A source of diffuse dust may also have been provided by an early generation of pre-galactic stars (ie. population III stars) associated with the formation of galactic haloes (McDowell 1986). Studies have shown that possible reddening from uniformly distributed IGM dust is limited by observations of radio-selected quasars. Since radio-selected quasars should have no bias against reddening by dust, such a component must be of sufficiently low optical depth to avoid producing a large fraction of `reddened' sources at high redshift (see Webster et al. 1995; Masci 1997).

In this paper, we show that a given quantity of dust has a much greater effect on the background universe when diffusely distributed. We shall investigate the effects of diffuse dust from, first, possible distributions on galaxy cluster scales and second, from a hypothesised uniformly distributed component in the IGM.

This paper is organised as follows: in the next section, we explore the dependence of background source counts observed through a given mass of dust on its spatial extent. In Section 3, we investigate the spatial distribution of dust optical depth through galaxy clusters and its effect on the counts and colours of background sources. Section 4 explores the consequences if all dust in the local universe were assumed uniformly distributed in the IGM. Further implications are discussed in Section 5 and all results are summarised in Section 6. All calculations use a Friedmann cosmology with and Hubble parameter where .

© Copyright Astronomical Society of Australia 1997