Revealing a hard X-ray spectral component that reverberates within one light hour of the central supermassive black hole in Ark 564

TL;DR

This study reveals a hard X-ray spectral component in Ark 564, likely caused by reprocessing of flaring emission by hot gas close to the supermassive black hole, using advanced timing and spectral analysis techniques.

Contribution

It identifies a new hard X-ray component in Ark 564 and links it to reprocessing by hot gas within one light hour of the black hole, using combined spectral and timing analysis.

Findings

Detection of delayed hard X-ray emission following soft X-ray flares

Reprocessing gas located at 10-100 gravitational radii from the black hole

Spectral hardening above 4 keV attributed to Compton-upscattering

Abstract

Ark 564 (z=0.0247) is an X-ray-bright NLS1. By using advanced X-ray timing techniques, an excess of "delayed" emission in the hard X-ray band (4-7.5 keV) following about 1000 seconds after "flaring" light in the soft X-ray band (0.4-1 keV) was recently detected. We report on the X-ray spectral analysis of eight XMM-Newton and one Suzaku observation of Ark 564. High-resolution spectroscopy was performed with the RGS in the soft X-ray band, while broad-band spectroscopy was performed with the EPIC-pn and XIS/PIN instruments. We analysed time-averaged, flux-selected, and time-resolved spectra. Despite the strong variability in flux during our observational campaign, the broad-band spectral shape of Ark 564 does not vary dramatically and can be reproduced either by a superposition of a power law and a blackbody emission or by a Comptonized power-law emission model. High-resolution spectroscopy revealed ionised gas along the line of sight at the systemic redshift of the source, with a low column density and a range of ionisation states. Broad-band spectroscopy revealed a very steep intrinsic continuum and a rather weak emission feature in the iron K band; modelling this feature with a reflection component requires highly ionised gas. A reflection-dominated or an absorption-dominated model are similarly able to well reproduce the time-averaged data from a statistical point of view, in both cases requiring contrived geometries and/or unlikely physical parameters. Finally, through time-resolved analysis we spectroscopically identified the "delayed" emission as a spectral hardening above ~4 keV; the most likely interpretation for this component is a reprocessing of the "flaring" light by gas located at 10-100 r_g from the central supermassive black hole that is so hot that it can Compton-upscatter the flaring intrinsic continuum emission.