Automated detection and modeling of binary microlensing events in OGLE-IV data. I. Events with well-separated bumps

Abstract

Gravitational microlensing depends primarily on the lens mass and presents a larger occurrence rate in crowded regions, which makes it the best tool to uncover the initial mass function (IMF) of low-mass stars in the Galactic bulge. The bulge IMF can be obtained from the luminosity function measured with the Hubble Space Telescope if one knows the statistics of binary stellar systems in the bulge. We aim to analyse a statistically significant number of binary-lens/single-source and single-lens/binary-source events, in order to explore the lower-mass end of the bulge IMF even in unresolved binary systems. This paper deals with events with clearly separated bumps and no caustic crossing or approach, whereas other types will be analysed in following works. A fully-automated approach in the search and modeling of binary events was implemented. Event detection was carried out with a modified version of the algorithm used in previous studies. Model fitting was carried out with Markov chain Monte Carlo and nested sampling methods, in order to find the most probable solution among binary lens or binary source models. We retrieved 107 binary events in Optical Gravitational Lensing Experiment (OGLE) light curves spanning ten years in 9 high-cadence and 112 low-cadence fields towards the bulge. Several criteria were applied to reduce false positives, resulting in 59 most likely binary lenses and 48 binary sources. The tools were effective in detecting a bona-fide sample of binary events, with a distribution of Einstein timescales around 35-40 days and flat distributions for mass ratio and source flux ratio. After proper consideration of detection efficiency, the statistics for binary fraction and mass ratio will provide valuable constraints for the bulge IMF.