Interpreting broad emission-line variations II: Tensions between luminosity, characteristic size and responsivity

TL;DR

This study models the variability of the broad Hβ emission line in AGN NGC 5548, examining how geometry, dust, and ionizing flux influence emission line properties and their observed delays.

Contribution

It introduces a detailed model linking BLR geometry, dust effects, and ionizing flux variations to observed emission line variability, highlighting the importance of a static outer boundary.

Findings

Predicted Hβ delays are too long in fixed boundary models.

Hβ variability amplitude matches observations except during low states.

Reducing incident ionizing flux can explain shorter observed delays.

Abstract

We investigate the variability behaviour of the broad Hb emission-line to driving continuum variations in the best-studied AGN NGC 5548. For a particular choice of BLR geometry, Hb surface emissivity based on photoionization models, and using a scaled version of the 13 yr optical continuum light curve as a proxy for the driving ionizing continuum, we explore several key factors that determine the broad emission line luminosity L, characteristic size R(RW), and variability amplitude (i.e., responsivity) eta, as well as the interplay between them. For fixed boundary models which extend as far as the hot-dust the predicted delays for Hb are on average too long. However, the predicted variability amplitude of Hb provides a remarkably good match to observations except during low continuum states. We suggest that the continuum flux variations which drive the redistribution in Hb surface emissivity F(r) do not on their own lead to large enough changes in R(RW) or eta(eff). We thus investigate dust-bounded BLRs for which the location of the effective outer boundary is modulated by the continuum level and the dust-sublimation and dust-condensation timescales. We find that in order to match the observed variability amplitude of broad Hb in NGC 5548 a rather static outer boundary is preferred. Intriguingly, we show that the most effective way of reducing the Hb delay, while preserving its responsivity and equivalent width, is to invoke a smaller value in the incident ionizing photon flux Phi(H) for a given ionizing source--cloud radial distance r, than is normally inferred from the observed UV continuum flux and typical models of the continuum SED.