2605.07965v1
A Changing-Look Seyfert Discovered by eROSITA Reveals a Two-Component Broad-Line Region
First listed 2026-05-08 | Last updated 2026-05-08
Abstract
Extreme sudden changes in the flow of accreting gas onto SMBHs manifest themselves via large-amplitude continuum variability and changes to broad Balmer emission profiles, driving changing-look AGN. X-ray flux monitoring with SRG/eROSITA revealed that in the Seyfert AGN HE 1237-2252 the soft X-ray flux dipped abruptly, by a factor of 17 within 18 months. We initiated a follow-up campaign that caught the luminosity recovery after the dip, and enabled us to study how the various accretion components responded during this flux recovery. Our campaign included multiband photometry, X-ray spectroscopy, and optical spectroscopy. We tracked as the accretion rate relative to Eddington increased by a factor of 7 in 3 years. Based on broad Hbeta variability, HE 1237-2252 was subtype 1.0-1.2 in 2002, transitioned to subtype 1.8 by the time of the luminosity dip, and then transitioned back to subtype 1.0 within 3 months as luminosity recovered. Both transitions saw broad Hbeta integrated line flux change by factors of 4-6. The broad Balmer profile is decomposed into a broad Gaussian consistent with virialized gas at 27+/-3 lt-dy, plus a double-peaked profile, consistent with a diskline structure at more than roughly 5 lt-dy. The diskline component's relative contribution to the total profile increases as continuum flux rises. The lack of obscuration in the X-ray spectra, as well as the IR continuum dip, point to an intrinsic pause in the accretion rate as opposed to variable line-of-sight obscuration. Candidates for the underlying mechanisms include propagating cold and warm fronts in the accretion disk. The increased prominence of the diskline BLR component's emission could be due to evolution in the physical extent of the X-ray corona, and in the fraction of >13.6 eV photons intercepted by the diskline, as the accretion rate increases.
Short digest
eROSITA caught HE 1237-2252 in a dramatic changing-look event, with the soft X-ray flux dropping by a factor of 17 in 18 months and a follow-up campaign tracking a roughly sevenfold rise in accretion rate relative to Eddington over the next 3 years. The source moved from Seyfert 1.0-1.2 in 2002 to 1.8 near the dip, then back to 1.0 within 3 months of recovery, while the broad Hbeta line flux changed by factors of 4-6. A key result is that the broad Balmer emission separates into two BLR components: a broad Gaussian from virialized gas at 27 +/- 3 light-days and a double-peaked diskline-like component arising beyond about 5 light-days, with the diskline contribution growing as the continuum brightens. Because the X-ray spectra show no strong obscuration and the IR continuum also dipped, the paper argues this was an intrinsic accretion pause, giving a rare view of how the BLR and corona reconfigure during a Seyfert changing-look cycle.
Key figures to inspect
- Figure 1 is the place to verify the event chronology: it ties the eROSITA/XMM/Swift/NICER continuum changes to the optical, UV, IR, and broad Balmer light curves, letting you see the factor-17 soft-X-ray drop, the recovery phase, and how quickly the line emission responded.
- Figure 2 should be inspected for the XMM spectral decomposition, especially to confirm the lack of strong obscuration and to see how the CompTT, hard power-law, and UxClumpy components divide up the X-ray continuum during the three XMM epochs.
- Figure 3 is useful for understanding the broadband accretion-flow interpretation: the fitted optical/UV/X-ray SEDs show how the disk, coronal, host-galaxy, and torus components shift between states and support the claim of a changing intrinsic accretion rate rather than a simple line-of-sight absorption event.
- Figure 4 gives the full sequence of 21 optical spectra, which is where the changing-look classification becomes visually obvious and where you can track the Balmer-profile evolution from a mainly broad Gaussian shape into a Gaussian plus increasingly prominent double-peaked structure.
Discussion
Log in to view the paper discussion, see votes, and leave your own feedback.