Peter Smith

I work on the internal dynamics of globular clusters, mostly by fitting dynamical models to diverse observational data in order to constrain the nature of the dark mass in these clusters.

The populations of stellar remnants and especially black holes that live in the dense cores of globular clusters have important implications for cluster evolution and as contributors to the gravitational wave signals that we now routinely detect. These remnants are difficult to observe directly, but their presence can be inferred through their effects on the overall dynamics of the cluster. I use dynamical models to fit the observed kinematics of globular clusters and infer the presence of dark mass in the cores of these clusters as well as characterize the nature of the dark mass.

In a recent paper, our group applied these dynamical modelling techniques to a large sample of Milky Way globular clusters, constraining their remnant fraction, black hole content, and several other properties. As part of this project, I led the validation work, which involved extracting realistic mock data from N-body models and then fitting the models to the mock data, ultimately demonstrating the effectiveness of the methodology.

Here we show the inferred black hole content of all 34 clusters in our sample.

My recent work has focused on using unconventional data, namely pulsar timing data, as additional constraints on the models. We have been able to show that, in clusters like 47 Tuc and Terzan 5, which have large populations of millisecond pulsars, the pulsar timing data can actually replace the conventional stellar kinematic data without compromising the precision of the methodology. Here we show the inferred black hole content of 47 Tuc and Terzan 5 for three subsets of the data. We show that for 47 Tuc, the pulsar timing data can replace the stellar kinematic data without compromising the precision of the methodology. For Terzan 5, where the stellar kinematic data is lacking, the pulsar timing data is essential for constraining the black hole content.

Here we show the inferred black hole content of 47 Tuc and Terzan 5 for three subsets of the data. We show that for 47 Tuc, the pulsar timing data can replace the stellar kinematic data without compromising the precision of the methodology. For Terzan 5, where the stellar kinematic data is lacking, the pulsar timing data is essential for constraining the black hole content.