Ten binary black hole mergers have already been detected during the first two observing runs of advanced LIGO and Virgo, and many more are expected to be observed in the near future. This opens the possibility for gravitational-wave (GW) astronomy to better constrain the properties of black hole binaries - not only as single sources, but as a whole astrophysical population. In this paper, we address the problem of using GW measurements to estimate the proportion of merging black holes produced either via isolated binaries or binaries evolving in young star clusters. To this end, we use a Bayesian hierarchical modeling approach applied to catalogs of merging binary black holes generated using state-of-the-art population synthesis and N-body codes. In particular, we show that, although current advanced LIGO/Virgo observations only mildly constrain the mixing fraction f ∈ [0, 1] between the two formation channels, we expect to narrow down the fractional errors on to f 10%-20% after a few hundreds of detections.
