Gravitational Waves as a Probe of Globular Cluster Formation and Evolution

TL;DR

Gravitational-wave signals from compact binary mergers can serve as powerful probes to understand the formation and evolution of globular clusters across cosmic history, especially with future advanced detectors.

Contribution

This paper demonstrates how future gravitational-wave detectors can trace globular cluster formation epochs, providing insights inaccessible to electromagnetic observations.

Findings

Peak cluster formation time can be measured within 0.05 Gyr after one year of data.

Formation epoch coinciding with universal star formation rate can be constrained within 0.4 to 10.5 Gyr.

Simulations show gravitational waves can reveal globular cluster formation history across redshifts.

Abstract

Globular clusters are considered to be likely breeding grounds for compact binary mergers. In this paper, we demonstrate how the gravitational-wave signals produced by compact object mergers can act as tracers of globular cluster formation and evolution. Globular cluster formation is a long-standing mystery in astrophysics, with multiple competing theories describing when and how globular clusters formed. The limited sensitivity of electromagnetic telescopes inhibits our ability to directly observe globular cluster formation. However, with future audio-band detectors sensitive out to redshifts of $z \approx 50$ for GW150914-like signals, gravitational-wave astronomy will enable us to probe the Universe when the first globular clusters formed. We simulate a population of binary black hole mergers from theoretically-motivated globular cluster formation models, and construct redshift measurements consistent with the predicted accuracy of third-generation detectors. We show that we can locate the peak time of a cluster formation epoch during reionisation to within 0.05Gyr after one year of observations. The peak of a formation epoch that coincides with the Universal star formation rate can be measured to within 0.4Gyr-10.5Gyr after one year of observations, depending on the relative weighting of the model components.