Invisible Monsters is a project funded by the Irish Research Council through a Government of Ireland Postdoctoral Fellowship awarded to Dr. Mark Kennedy.

Abstract:
Every star in our Galaxy will end its life as one of three different objects - a black hole, a neutron star, or a white dwarf. While there are hundreds of neutron stars and thousands of white dwarfs known within our galaxy, there are only 18 strong black hole candidates. This is in direct conflict with simulations which suggest there should be over 100 million black holes in the Milky Way. So where are these missing black holes? The same simulations suggest that at least 1 million of them should be in binary star systems, but since black holes do not emit any radiation, detecting them has proven to be difficult. However, recent results suggest we are on the precipice of being able to find them due to the advent of powerful, next generation all-sky surveys. These surveys will be able to detect small variations in light coming from the companion stars caused by the black holes in these binaries. Throughout the course of this fellowship, I will use recently developed methods in machine learning to identify candidate black holes in these systems based on these small variations in light. Candidates will then be followed up with larger telescopes and modelled to measure the masses of each of the binary components, confirming the mass of the invisible primaries inthese binaries. Through this project, we aim to increase the population of known black holes by at least 50%, while also establishing a robust pipeline for identifying further candidates in the future as more sensitive survey telescopes are built.

Research Plans
The current focus of this project is to identify suitable binaries which may contain black holes as their primary star. We have combined the data from ZTF and GAIA to produce light curves and positions on the H-R diagram of ~10,000 potential systems, along with folded light curves for the strongest peaks in their periodograms (see attached figure for an example of our dashboard). The most promising systems are currently being modelled with Icarus (Breton et al. 2012, Link: https://ui.adsabs.harvard.edu/abs/2012ApJ...748..115B/abstract; Kennedy et al. 2022, Link: https://ui.adsabs.harvard.edu/abs/2022MNRAS.512.3001K/abstract). Once this modelling has been completed, proposals will be prepared to follow up the most promising candidates.