The UV-bright, Slowly Declining Transient PS1-11af as a Partial Tidal Disruption Event

TL;DR

The paper reports the discovery and analysis of PS1-11af, a long-lasting, blue transient event likely caused by a partial tidal disruption of a star by a supermassive black hole, distinguished from supernovae and other models.

Contribution

It provides the first detailed analysis of PS1-11af as a partial tidal disruption event, with multi-wavelength observations and modeling that differentiate it from supernovae and full disruptions.

Findings

PS1-11af is not consistent with supernova properties.

The event's luminosity suggests a partial disruption with only ~0.002 solar masses accreted.

No relativistic outflows detected in deep radio observations.

Abstract

We present the Pan-STARRS1 discovery of the long-lived and blue transient PS1-11af, which was also detected by GALEX with coordinated observations in the near-ultraviolet (NUV) band. PS1-11af is associated with the nucleus of an early-type galaxy at redshift z=0.4046 that exhibits no evidence for star formation or AGN activity. Four epochs of spectroscopy reveal a pair of transient broad absorption features in the UV on otherwise featureless spectra. Despite the superficial similarity of these features to P-Cygni absorptions of supernovae (SNe), we conclude that PS1-11af is not consistent with the properties of known types of SNe. Blackbody fits to the spectral energy distribution are inconsistent with the cooling, expanding ejecta of a SN, and the velocities of the absorption features are too high to represent material in homologous expansion near a SN photosphere. However, the constant blue colors and slow evolution of the luminosity are similar to previous optically-selected tidal disruption events (TDEs). The shape of the optical light curve is consistent with models for TDEs, but the minimum accreted mass necessary to power the observed luminosity is only ~0.002M_sun, which points to a partial disruption model. A full disruption model predicts higher bolometric luminosities, which would require most of the radiation to be emitted in a separate component at high energies where we lack observations. In addition, the observed temperature is lower than that predicted by pure accretion disk models for TDEs and requires reprocessing to a constant, lower temperature. Three deep non-detections in the radio with the VLA over the first two years after the event set strict limits on the production of any relativistic outflow comparable to Swift J1644+57, even if off-axis.