Evolution of the ionizing background at high redshifts

Abstract

The decrease in number density of Lyman-α clouds near the background quasar is an observational result which is often called the 'proximity' or 'inverse' effect. It is thought that, for nearby clouds, the quasar's flux dominates the background radiation field, increasing the ionization state of the clouds and reducing the (observed) H I column density.In this paper we analyse a sample of 11 quasars from the literature for which accurate column density estimates of the Lyman-α lines exist. We confirm, to a significance level of more than 3 standard deviations, that the proximity effect exists. If it is related to the background flux then the intensity and evolution of the background have been constrained.Using a maximum likelihood method, we determine the strength of the extragalactic ionizing background for 2.0 < z < 4.5, taking account of possible systematic errors in our determination and estimating the effect of biases inherent in the data. If the background is constant we find that it has an intensity of 100 +50 −30 J 23 , where J 23 is defined as 10 −23 ergs cm −2 Hz −1 sr −1 . There is no significant evidence for a change in this value with redshift.

Faint Active Galactic Nuclei and the Ionizing Background

“…An estimate of the ionizing rate can be obtained by determining the distance from the quasar at which the number of lines is equal to the background expectation. Over the years, measurements of J À21 (assuming UV $ 1:8) have taken values between $0.7 and $3 (Williger et al 1994;Bechtold 1994;Fernandez-Soto et al 1995;Cristiani et al 1995;Giallongo et al 1996;Cooke, Espey, & Carswell 1997;Scott et al 2000;Liske & Williger 2001). Since what is effectively measured is a ratio of proper distances, this analysis should be relatively independent of cosmological parameters.…”

“…This is because QSOs are likely to be found in environments that are denser than average (Pascarelle et al 2001;Ellison et al 2002), so that the regions of excess ionization end up being smaller than they would have been if the region was of average density. Such overdensities would likely scale with the mass/ luminosity of the AGN, which might explain the slight anticorrelation between more luminous QSOs and their proximity effects (see Cooke et al 1997). claim that this bias could cause the proximity effect measurements to overestimate the background by as much as a factor of 3.…”

An Ionizing Ultraviolet Background Dominated by Massive Stars

“…QSOs and star-forming galaxies have long been considered the two prime candidates. It has been argued that QSOs may fall short of producing a metagalactic flux as high as that inferred from the proximity effect at high redshift (Bechtold 1994;Giallongo et al 1996;Cooke, Espey, & Carswell 1997;Scott et al 2000) since their space density declines rapidly at early epochs (Shapiro & Giroux 1987;Miralda-Escudé & Ostriker 1990;Meiksin & Madau 1993;Madau, Haardt, & Rees 1999). The detection of a numerous population of Lyman break galaxies (LBGs) at (Steidel et al 1996) makes the idea of massive z ≈ 3 stars dominating the UV background at early epochs quite plausible.…”

Faint AGN and the Ionizing Background

“…An estimate of the ionizing rate can be obtained by determining the distance from the quasar at which the number of lines is equal to the background expectation. Over the years, measurements of J −21 (assuming α UV ∼ 1.8) have taken values between ∼ 0.7 and ∼ 3 (Williger et al 1994;Bechtold 1994;Fernández-Soto et al 1995;Cristiani et al 1995;Giallongo et al 1996;Cooke, Espey & Carswell 1997;Scott et al 2000;Liske & Williger 2001). Since what is effectively measured is a ratio of proper distances, this analysis should be relatively independent of cosmological parameters.…”

“…This is because QSOs are likely to be found in environments that are denser than average (Pascarelle et al 2001;Ellison et al 2002), so that the regions of excess ionization end up being smaller than they would have been if the region was of average density. Such overdensities would likely scale with the mass/luminosity of the AGN, which might explain the slight anti-correlation between more luminous QSOs and their proximity effects (see Cooke, Espey & Carswell 1997). claim this bias could cause the proximity effect measurements to overestimate the background by as much as a factor of 3.…”