Automated transient detection in the context of the 4m ILMT

Abstract

In the era of sky surveys like Palomar Transient Factory (PTF), Zwicky Transient Facility (ZTF) and the upcoming Vera Rubin Observatory (VRO) and ILMT, a plethora of image data will be available. ZTF scans the sky with a field of view of 48 deg${}^2$ and VRO will have a FoV of 9.6 deg${}^2$ but with a much larger aperture. The 4m ILMT covers a 22${}^{\prime }$ wide strip of the sky. Being a zenith telescope, ILMT has several advantages like low observation air mass, best image quality, minimum light pollution and no pointing time loss. Transient detection requires all these imaging data to be processed through a Difference Imaging Algorithm (DIA) followed by subsequent identification and classification of transients. The ILMT is also expected to discover several known and unknown astrophysical objects including transients. Here, we propose a pipeline with an image subtraction algorithm and a convolutional neural network (CNN) based automated transient discovery and classification system. The pipeline was tested on ILMT data and the transients as well as variable candidates were recovered and classified.

9 pages, 3 figures, accepted for publication in the Bulletin of Liège Royal Society of Sciences as a part of 3rd Belgo-Indian Network for Astronomy and Astrophysics (BINA) workshop, 22-24 March 2023