GALEX colours of quasars and intergalactic medium opacity at low redshift

TL;DR

This study uses GALEX UV colours of quasars to estimate intergalactic medium opacity at low redshift, revealing higher Lyman limit system densities than spectroscopic methods and suggesting a tension with current quasar ionising spectra.

Contribution

It introduces a novel UV colour-based approach combined with Monte Carlo simulations to quantify IGM opacity at low redshift, challenging existing density estimates.

Findings

Higher LLS density than spectroscopic counts.

Robust results despite colour dispersion uncertainties.

Tension between IGM opacity models and quasar ionising spectra.

Abstract

The distribution of neutral hydrogen in the intergalactic medium (IGM) is currently explored at low-z by means of UV spectroscopy of quasars. We propose an alternative approach based on UV colours of quasars as observed from GALEX surveys. We built a NUV-selected sample of 9033 quasars with (FUV-NUV) colours. The imprint of HI absorption in the observed colours is suggested qualitatively by their distribution as a function of quasar redshift. Because broad band fluxes lack spectral resolution and are sensitive to a large range of N_HI a Monte Carlo simulation of IGM opacity is required for quantitative analysis. It was performed with absorbers randomly distributed along redshift and column density distributions, assumed to be a broken power law with index beta1 (10^15 < N_HI <10^17.2 cm^-2) and beta2 (10^17.2 < N_HI <10^19 cm^-2). The redshift distribution is proportional to the redshift evolution law of the number density of Lyman limit systems (LLS) per unit redshift as determined by spectroscopic surveys.The simulation is run with different assumptions on the spectral index alpha_nu of the quasar ionising flux. The fits between the simulated and observed distribution of colours require an LLS redshift density larger than that derived from spectroscopic counting. This result is robust in spite of difficulties in determining the colour dispersion other than that due to HI absorption. We provide arguments to retain alpha_nu = - 2, a value already extreme with respect to those measured with HST/COS. Further fitting of power law index beta1 and beta2 leads to a higher density by a factor of 1.7 (beta1 = -1.7, beta2 = -1.5), possibly 1.5 (beta1 = -1.7, beta2 = -1.7). Beyond the result in terms of density the analysis of UV colours of quasars reveals a tension between the current description of IGM opacity at low z and the published average ionising spectrum of quasars.