Merging binary black holes, the first ten years

The black-hole mergers observed by LIGO and Virgo are ultimately driven by the emission of energy and angular momentum via gravitational waves. Yet general relativity alone is not enough to explain the existence of merging compact-object binaries. For instance, black-hole binaries of ~10 M☉ orbiting at separations of ~10 R☉ would take longer than a Hubble time to merge under gravitational radiation reaction alone.

Additional astrophysical processes are therefore required to bring these binaries into the gravitational-wave regime. Understanding the origin and evolution of merging compact binaries remains one of the central challenges in gravitational-wave astronomy and has motivated the development of a variety of formation scenarios. In this talk, I will present a status update on the formation-channel problem, with a specific focus on hierarchical black-hole mergers.

Along the way, I will highlight the statistical challenge we face: inferring the properties of a population of sources from noisy measurements subject to strong selection effects. Recent advances in hierarchical Bayesian inference--and, why not, a touch of machine learning--are enabling us to extract increasingly detailed information from gravitational-wave data. These methods are laying the foundation for the time when our field will fully transition into a genuine big-data science.