The properties of GSN 069 accretion disk from a joint X-ray and UV spectral analysis: stress-testing quasi-periodic eruption models

TL;DR

This study analyzes GSN 069's accretion disk using joint X-ray and UV data, revealing a viscously expanding disk post-TDE that challenges current QPE models.

Contribution

It provides the first detailed UV and X-ray spectral analysis of GSN 069, constraining disk properties and testing QPE formation models.

Findings

The disk expanded from 2014 to 2018, showing cooling and growth.

Disk inclination angle constrained between 30° and 65°.

Current models cannot fully explain the observed disk stability and QPE occurrence.

Abstract

We present an analysis of Hubble Space Telescope (HST) and XMM-Newton data of the tidal disruption event (TDE) candidate and quasi-periodic eruption (QPE) source GSN 069. Using ultraviolet (UV) and optical images at HST resolution, we show that GSN 069's emission consists of a point source superimposed on a diffuse stellar component. The latter accounts for $\leq 5\%$ of the UV emission in the inner 0.5"$\times$0.5" region, while the luminosity of the former cannot be attributed to stars. Analyzing the 2014/2018 \hst UV spectra, we show that to leading order the intrinsic spectral shape is $\nu\,L_{\nu}\propto\nu^{4/3}$, with $\sim10\%$ far UV flux variability between epochs. The contemporaneous X-ray and UV spectra can be modeled self-consistently in a thin disk framework. At observed epochs, the disk had an outer radius ($R_{\rm out}$) of $\mathcal{O}(10^3R_{\rm g})$, showing both cooling and expansion over four years. Incorporating relativistic effects via numerical ray tracing, we constrain the disk inclination angle ($i$) to be $30^\circ\,\lesssim\,i\,\lesssim\,65^\circ$ and identify a narrow region of spin-inclination parameter space that describes the observations. These findings confirm that GSN 069 hosts a compact, viscously expanding accretion disk likely formed after a TDE. Implications for QPE models are: (i) No published disk instability model can explain the disk's stability in 2014 (no QPEs) and its instability in 2018 (QPEs present); (ii) While the disk size in 2018 allows for orbiter/disk interactions to produce QPEs, in 2014 the disk was already sufficiently extended, yet no QPEs were present. These findings pose challenges to existing QPE models.