Tidal disruption events: what, where, when and why?

Abstract

A tidal disruption event (TDE) occurs when a star crosses the tidal radius of a nearby supermassive black hole (SMBH) and is subsequently torn apart. The fall-back of the stellar debris onto the black hole produces a luminous flare which we can use to understand the black hole and its environment. In particular TDEs allow us to probe SMBHs in otherwise quiescent galaxies and gain insight into the lower mass SMBH population distribution. The fact that TDEs evolve on human timescales make them excellent tools with which to probe accretion physics, allowing us to observe the formation of accretion disks and jets. Over the past decade, with the introduction of high cadence all-sky surveys an increasingly large number of TDEs with well sampled lightcurves and spectra have been discovered. These events have shown a remarkably diverse range of observational features and a strange preference to occur in E+A, or post-starburst galaxies. These two factors have given rise to two key questions: ’What processes produce the observed emission in TDEs?’ and ’Why is it that TDEs occur where and when that they do?’.

In this thesis I present three separate studies which aim at answering these questions. The first project is a detailed spectroscopic analysis of the TDE AT2018hyz during outburst. The dataset used is a combination of spectra from a number of telescopes resulting in high-cadence spectroscopic follow-up. This allowed me to monitor closely the evolution of spectroscopic features. The Balmer and He II emission line profiles varied dramatically throughout the TDE’s evolution. Most notably from ∼ 30 − 100 days after optical lightcurve peak the Balmer lines develop a double-peaked shape. This, combined with an unusually low Hα/Hβ ratio lead me to determine that the observations show the formation of an accretion disk before it either dissipates or gets obscured from view. At the time this was the most clear sign of accretion disk formation in a TDE from optical spectra.

The second project is my analysis of another TDE, AT2019qiz. Here I present and analyse very late time optical and X-ray spectra. The optical spectra show high ionisation coronal emission lines which have never before been seen in a TDE and a jump in the flux of narrow lower ionisation lines. The X-ray spectrum shows an order-of-magnitude jump in soft X-ray flux compared to a previous observation taken closer to the peak of the TDE outburst. The coronal line emission implies a source of highly ionising extreme UV photons which is predicted by accretion models but until now not shown in observations. From the emission line widths I show that the coronal lines originate from a region further out from the SMBH than the broad line region, but within the narrow line region distance. In addition, the distances imply that the structure reprocessing the material must have already been in place at the time of outburst, likely originating from previous nuclear activity. The finding raises questions as to why these lines appear in AT2019qiz and not other TDEs and I recommend systematically searching for them in late time spectra of other TDEs. In the final project I analyse spectra of a sample of 11 host galaxies of tidal disruption events. By measuring spectral indices of the galaxies such as D n 4000 and Lick Hδ A , I find that 6 of the objects fit the literature definition of a post- starburst galaxy. This is backed up by comparisons of my measurements to galaxy models from the Bagpipes synthetic population synthesis code. In the second half of the project I conduct detailed fitting to my host galaxy spectra using Bagpipes with the goal of uncovering the star formation histories. I find that the objects previously defined as post-starburst have had a burst of star-formation within the last few Gyrs, and that one object with ongoing star-formation may be undergoing a burst at the time of TDE outburst. In general my results are consistent with other’s findings that the TDE host-galaxy population is over- represented by E+A galaxies, and that this may be related to the presence of nuclear star clusters in TDE hosts.