Continuum Reverberation in Bright Quasars Using NASA/ATLAS

Abstract

Variable continuum emission from AGN can be used to probe the structure of their accretion disks via reverberation mapping. Assuming a variable, hot inner light source irradiates the surrounding accretion disk, time delays between different continuum band light curves reveal light-travel times between their respective emission regions. Inter-band delays measured in several low-luminosity AGN are ubiquitously $\sim 3$ times longer than expected from standard disk theory, with evidence this size discrepancy may decrease in more luminous AGN. We have analysed high-cadence light curves of 9,498 of the brightest quasars between redshift 0.3-2.5 in the largest continuum reverberation study to date. Given the large sample size, we construct bins and fit delays jointly to combine inference across the parameter space and improve lag detections. We find that the size discrepancy persists in our high-luminosity sample, and that the previously seen anti-correlation with luminosity is likely driven by wavelength effects. The complex, non-monotonic wavelength dependence of delay amplitudes strongly suggests that contamination of inter-band delays by variable diffuse emission is widespread in the AGN population. We test delay behaviour against a variety of quasar properties finding longer lags in quasars with: higher Eddington ratios, redder colours, stronger optical FeII equivalent widths, higher iron ratios (both UV FeII/MgII and optical FeII/H$β$), CIV broad absorption troughs, and lower CIV blueshift.

Figures (22)

• Figure 1: Parent sample: luminosities and black hole masses vs. redshift compared to the wider SDSS quasar population illustrating our chosen sample is far among the most luminous. Bolometric luminosities are estimated from the 2020_rakshit_spectral$\lambda L_{3000}$ values using a bolometric correction of 5.15 2006_richards_spectral. A weaker correlation of black hole mass with both quantities is present.
• Figure 2: Left: ATLAS colour dependence on redshift. Right: Distribution of colour residuals standardised by redshift-dependent median and MAD. The median ($\pm3$MAD) is shown as the solid (dashed) black line in both panels. Grey points and grey distribution tails represent removed objects.
• Figure 3: Redshift evolution of Feii EW and its ratio to the relevant low-ionisation line in each regime (separated by the black dashed line $z=0.72$). A running median and MAD trend is displayed in grey. The marginal distributions for each regime are given in the right-hand panels, along with the distribution medians.
• Figure 4: Civ blueshift anti-correlation with equivalent width for the non-BAL quasars with sufficient line SNR, luminosity and FWHM. Black dashed lines represent sample cuts for equivalent width and blueshift. Red dashed line represent the $\leq-1800\,{\rm km\,s^{-1}}$ wind velocity cut-off.
• Figure 5: Time delay seen by a distant observer along the $z$-axis for an X-ray photon reprocessed by the disk where $i$ is the system inclination and $R$ ($\theta$) is the radial (azimuthal) position.