Poster abstract details

X-ray transient classification for eROSITA
Adam Malyali, Arne Rau, Kirpal Nandra


eROSITA on-board the SRG satellite will perform the next large X-ray all-sky survey. With its 30-fold increased sensitivity relative to its predecessor ROSAT and its multi-visit, multi-cadence survey strategy, eROSITA will provide a new and deeper look into X-ray time domain astrophysics. This holds the potential for the discovery of new populations of exotic, interesting and unexpected transient sources. However, these must first be detected amongst the millions of AGN, galaxy clusters and stars that will dominate the detected objects. Based on end-to-end simulations, we present the expected detection rate of white dwarf - black hole tidal disruption events with eROSITA. Furthermore, we discuss the challenges of developing machine learning algorithms for classification of X-ray transients for eROSITA, where the only training datasets available are non-representative and biased; as well as the results of an unsupervised approach to transient classification and outlier detection.