Master's thesis defence by Ivan Stoyanov Kanev

Broad Emission Lines as a Probe of the Ionising SED in Mrk 590

A Cloudy-Based Photoionisation Study of a Changing-Look AGN 

Abstract

Mrk 590 is one of the most studied changing-look active galactic nuclei (AGN). It faded toward a Seyfert 2 galaxy state by the early 2010s and re-brightened from 2017 onward. Joint X-ray and UV modelling by Lawther et al. (2025) attributes this re-brightening to an intrinsic accretion-state change.

To investigate the same transition from the side of the broad-line region (BLR) gas, I decompose two X-Shooter optical spectra of Mrk 590 in two flux states. The low flux state spectrum is from 2022 and the high flux state spectrum is from 2025. I decompose the spectra with a spectrum fitting pipeline I wrote for this thesis. From the resulting fits I measure the broad-line equivalent widths (EWs) ofbHα and Hβ at each spectrum. I then generate Cloudy photoionisation grids in the (log nH, logΦH) plane, one for each of the two ionising-SED from Lawther et al. (2025) and locate the observed EWs on the predicted-EW contour maps by assuming a single cloud model.

Both ionising SEDs reproduce the observed EWs. Both broad-line EWs decrease from the low- to the high-flux state by factors of ∼ 2.5 (Hα) and ∼ 4 (Hβ), even though the continuum brightens by a factor of ∼ 24 at Hβ. Meaning, that the BLR reprocesses a smaller fraction of the incident continuum into line emission. A single measured EW selects a contour, so one emission line cannot pin the gas density.

The requirement that Hα and Hβ have to be reproduced from the same gas narrows the allowed region on the (log nH, logΦH) grids. Under both ionising SEDs a single cloud at one density then reproduces the EWs of both lines in both spectra, with only the incident ionising photon flux (logΦH) rising between the two states. My solution places the cloud at log nH = 12, with logΦH rising from logΦH = 19.5 in 2022 to logΦH = 20.2 in 2025.

Densities below log nH = 10 are ruled out, as the flux required to match the high-state EW would over-ionise the gas, causing the emission line to disappear. The permitted-narrow line components increase in width three fold as the luminosity from the central engine increases. Additionally, a blueshifted intermediate-width component and a coronal line that appear only in the high state, independently points to a physical rather than obscurationdriven change.

The increase in ionising flux at fixed density is the signature of an intrinsic accretionstate change. Meaning, the optical broad lines give a gas-based witness to the conclusion of Lawther et al. (2025), i.e. that the central engine has reactivated between the two flux state spectra. My analysis is a consistency check, not a fit to the SED parameters. At the unit covering fraction adopted for the grids the absolute BLR conditions are degenerate and only the relative result is robust to that degeneracy.

Supervisor

Marianne Vestergaard

External examiner

Jérôme Chenevez (DTU)