No sign of G2's encounter affecting Sgr A*'s X-ray flaring rate from $Chandra$ observations

TL;DR

This study analyzes extensive Chandra X-ray data to determine if G2's close approach to Sgr A* affected its X-ray flaring rate, finding no significant change and challenging previous claims of increased activity.

Contribution

The paper provides a careful, contamination-free analysis of Sgr A*'s X-ray flaring rate using Chandra data, showing no evidence of change post-G2 encounter, unlike prior reports.

Findings

No significant increase in flaring rate after G2's pericenter passage

Analysis accounts for contamination and instrument differences

Results are consistent with a constant flaring rate hypothesis

Abstract

An unusual object, G2, had its pericenter passage around Sgr A*, the $4\times10^6$ M$_\odot$ supermassive black hole in the Galactic Centre, in Summer 2014. Several research teams have reported evidence that following G2's pericenter encounter the rate of Sgr A*'s bright X-ray flares increased significantly. Our analysis carefully treats varying flux contamination from a nearby magnetic neutron star and is free from complications induced by using data from multiple X-ray observatories with different spatial resolutions. We test the scenario of an increased bright X-ray flaring rate using a massive dataset from the \textit{Chandra X-ray Observatory}, the only X-ray instrument that can spatially distinguish between Sgr A* and the nearby Galactic Centre magnetar throughout the full extended period encompassing G2's encounter with Sgr A*. We use X-ray data from the 3 Ms observations of the \textit{Chandra} \textit{X-ray Visionary Program} (XVP) in 2012 as well as an additional 1.5 Ms of observations up to 2018. We use detected flares to make distributions of flare properties. Using simulations of X-ray flares accounting for important factors such as the different $Chandra$ instrument modes, we test the null hypothesis on Sgr A*'s bright (or any flare category) X-ray flaring rate around different potential change points. In contrast to previous studies, our results are consistent with the null hypothesis; the same model parameters produce distributions consistent with the observed ones around any plausible change point.