Systematic collapse of the accretion disc in AGN confirmed by UV photometry and broad line spectra

TL;DR

This study confirms that in low Eddington ratio AGN, the accretion disc evaporates, leading to a significant drop in UV emission and the disappearance of broad emission lines, indicating a population of UV-faint, X-ray bright AGN.

Contribution

It provides observational evidence of the disc evaporation in AGN at low accretion rates, linking UV faintness with the absence of broad emission lines, confirmed through UV photometry and spectral analysis.

Findings

UV emission drops significantly below L/L_Edd ~ 0.01

Broad Hβ emission line disappears at low accretion rates

Identifies a population of UV-faint, X-ray bright AGN lacking broad lines

Abstract

A recent study on the spectral energy distribution (SED) of AGN combined unobscured X-ray sources from the eROSITA eFEDS Survey with high quality optical imaging from Subaru's Hyper Suprime-Cam (HSC). The HSC data enabled accurate host galaxy subtraction as well as giving a uniform black hole mass estimator from the stellar mass. The resulting stacked optical/X-ray SEDs for black holes at fixed mass show a dramatic transition, where the dominating disc component in bright AGN evaporates into an X-ray hot plasma below $L/L_{\rm Edd}\sim 0.01$. The models fit to these datasets predicted the largest change in SED in the rest frame UV ($< 3000\,\r{A}$), but this waveband was not included in the original study. Here we use archival $u$-band and UV photometry to extend the SEDs into this range, and confirm the UV is indeed intrinsically faint in AGN below $L/L_{\rm Edd}\sim 0.01$ as predicted. This dramatic drop in UV photo-ionising flux is also seen from its effect on the broad emission lines. We stack the recently released SDSS DR18 optical spectra for this sample, and show that the broad H$\beta$ line disappears along with the UV bright component at $L/L_{\rm Edd}\sim 0.01$. This shows that there is a population of unobscured, X-ray bright, UV faint AGN which lack broad emission lines (true type 2 Seyferts).