Using AGN lightcurves to map accretion disc temperature fluctuations

TL;DR

This paper presents a new model for AGN variability, showing that observed UV-optical lightcurves are dominated by slow ingoing waves triggering central fluctuations, challenging the traditional lamppost interpretation.

Contribution

It introduces a linearized accretion disc model fitting UV-optical lightcurves to reveal dominant slow ingoing wave patterns, contrasting with the outgoing wave assumption in lamppost models.

Findings

Most AGN fluctuations are dominated by slow ingoing waves.

Lightcurves are primarily influenced by central temperature fluctuations.

Slower waves' contributions are exponentially smoothed, affecting variability interpretation.

Abstract

We introduce a new model for understanding AGN continuum variability. We start from a Shakura--Sunyaev thin accretion disc with a steady-state radial temperature profile $T(R)$ and assume that the variable flux is due to axisymmetric temperature perturbations $\delta T(R,t)$. After linearizing the equations, we fit UV-optical AGN lightcurves to determine $\delta T(R,t)$ for a sample of seven AGNs. We see a diversity of $|\delta T/T| \sim 0.1$ fluctuation patterns which are not dominated by outgoing waves traveling at the speed of light as expected for the "lamppost" model used to interpret disc reverberation mapping studies. Rather, the most common pattern resembles slow ($v \ll c$) ingoing waves. An explanation for our findings is that these ingoing waves trigger central temperature fluctuations that act as a lamppost, producing lower amplitude temperature fluctuations moving outwards at the speed of light. The lightcurves are dominated by the lamppost signal -- even though the temperature fluctuations are dominated by other structures with similar variability time-scales -- because the discs exponentially smooth the contributions from the slower moving ($v \ll c$) fluctuations to the observed lightcurves. This leads to lightcurves that closely resemble the expectations for a lamppost model but with the slow variability time-scales of the ingoing waves. This also implies that longer time-scale variability signals will increasingly diverge from lamppost models because the smoothing of slower-moving waves steadily decreases as their period or spatial wavelength increases.