Predicting the self-lensing population in optical surveys.

The vast majority of binaries containing a compact object and a regular star spend most of their time in a quiescent state where no strong interactions occur between components. Detection of these binaries is extremely challenging and only few candidates have been detected through optical spectroscopy. Self-lensing represents a new means of detecting compact objects in binaries, where gravitational lensing of the light from the visible component by the compact object produces periodic optical flares. Here we show that current and planned large-area optical surveys can detect a significant number ($\sim 100$-$10,000$s) of these self-lensing binaries and provide insights into the properties of the compact lenses. We show that many of the predicted population of observable self-lensing binaries will be observed with multiple self-lensing flares; this both improves the chances of detection and also immediately distinguishes them from chance-alignment micro-lensing events. Through self-lensing we can investigate long - but previously hidden - stages of binary evolution and consequently provide new constraints on evolutionary models which impact on the number and nature of double compact object mergers.