2020MNRAS.494.5917F

The great power offered by photoionization models of active galactic nuclei emission line regions has long been mitigated by the fact that very little is known about the spectral energy distribution (SED) between the Lyman limit, where intervening absorption becomes a problem, and 0.3 keV, where soft X-ray observations become possible. The emission lines themselves can, to some degree, be used to probe the SED, but only in the broadest terms. This paper employs a new generation of theoretical SEDs that are internally self-consistent, energy conserving, and tested against observations, to infer properties of the emission-line regions. The SEDs are given as a function of the Eddington ratio, allowing emission-line correlations to be investigated on a fundamental basis. We apply the simplest possible tests, based on the foundations of photoionization theory, to investigate the implications for the geometry of the emission-line region. The SEDs become more far-ultraviolet bright as the Eddington ratio increases, so the equivalent widths of recombination lines should also become larger, an effect that we quantify. The observed lack of correlation between Eddington ratio and equivalent width shows that the cloud covering factor must decrease as Eddington ratio increases. This would be consistent with recent models proposing that the broad-line region is a failed dusty wind off the accretion disc.